Radiation-associated sarcomas

INTRODUCTION — Improvements in therapy for a variety of malignancies have led to increasing numbers of people who are long-term cancer survivors. Survivors of both adult and childhood cancers are at risk for developing therapy-related complications, including second cancers, now aptly termed an "iatrogenic disease of success" [1]. (See "Overview of cancer survivorship care for primary care and oncology providers", section on 'Epidemiology'.)

Therapeutic radiation has long been recognized as an inducing agent in the development of malignant neoplasms [2-5]. The first case reports of bone sarcomas in patients who had received radiation therapy (RT) for benign bone conditions were published in 1922, and a report of bone sarcomas in the jaws of radium-dial painters followed in 1929 [6,7].

Sarcomas are rare malignant tumors that arise from mesenchymal tissues at any location. The histopathologic spectrum of sarcomas is broad, reflecting the fact that the embryonic mesenchymal cells from which they arise have the capacity to mature into striated skeletal and smooth muscle, adipose and fibrous tissue, bone, and cartilage. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Histopathology'.)

Besides exposure to RT, pathogenetic factors include genetic predisposition, exposure to chemotherapy, and for certain types of soft tissue sarcoma, chronic edema, and viral infection. (See "Pathogenetic factors in soft tissue and bone sarcomas".)

This topic review will cover secondary soft tissue and bone sarcomas that arise after therapeutic irradiation in adults and children. Other radiation-associated malignancies (eg, thyroid cancer, acute and chronic myelogenous leukemia, breast and lung cancer, and secondary malignancies following hematopoietic cell transplantation) as well as an overview of the consequences of unintended radiation exposure, and the risk of malignancy due to diagnostic imaging are covered separately:

●(See "Acute myeloid leukemia: Pathogenesis", section on 'Ionizing radiation' and "Factors that modify breast cancer risk in women", section on 'Exposure to therapeutic ionizing radiation'.)

●(See "Clinical manifestations and diagnosis of chronic myeloid leukemia", section on 'Epidemiology'.)

●(See "Radiation-induced thyroid disease".)

●(See "Cigarette smoking and other possible risk factors for lung cancer", section on 'Radiation therapy'.)

●(See "Factors that modify breast cancer risk in women", section on 'Exposure to diagnostic radiation'.)

●(See "Secondary cancers after hematopoietic cell transplantation".)

●(See "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure".)

●(See "Clinical manifestations, evaluation, and diagnosis of acute radiation exposure".)

●(See "Radiation-related risks of imaging".)

●(See "Radiation dose and risk of malignancy from cardiovascular imaging".)

EPIDEMIOLOGY AND HISTOLOGIC DISTRIBUTION — Although the use of radiation therapy (RT) increases the risk of sarcoma compared with no irradiation, the absolute risk is low overall. Current estimates suggest that RT-associated sarcomas account for only 3 to 6 percent of all sarcomas [8-16].

The incidence of an RT-associated sarcoma has been estimated to be well under 1 percent in exposed adult populations [8]:

●In a series of patients receiving RT for a variety of different malignancies, the incidence of a radiation-associated sarcoma was 0.11 percent (9 of 7772 patients) after orthovoltage radiation (kilovoltage, low-energy, superficial radiographs used for treating superficial structures) and 0.09 percent (13 of 14,534) after megavoltage radiation (higher-energy deep radiation delivered via modern linear accelerator or a Cobalt-60 unit that can treat deep-seated tumors) [17].

●Although women who undergo RT for breast cancer have an increased risk of in-field sarcomas that persist for 20 to 30 years or longer, the absolute magnitude of the risk appears to be small (table 1). (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Radiation exposure in breast cancer survivors' and "Overview of long-term complications of therapy in breast cancer survivors and patterns of relapse", section on 'Risks associated with radiation therapy'.)

●Among women undergoing RT for a primary gynecologic tumor, the absolute risk of a postradiation sarcoma is estimated to be between 0.03 to 0.8 percent [18].

●In a single institution study, 53 of the 39,118 patients who received definitive RT for nasopharyngeal cancer between 1964 and 2003 were diagnosed with a radiation-associated sarcoma (crude incidence 0.14 percent) [19].

Children are more susceptible to radiation-associated sarcoma than are adults, perhaps for reasons of biology and not just their longer follow-up time after treatment [12-14,20-25]:

●Nguyen reported an incidence of 1.2 percent (36 bone and 16 connective tissue, from 124 secondary malignancies in 4401 patients) in a pediatric population treated with RT for a variety of malignancies [20].

●In a European pediatric cohort of 4581 patients treated for a solid cancer during childhood, 1.3 percent developed secondary sarcomas (37 bone and 24 connective tissue), which represented 42 percent of all secondary malignancies [21].

●As an example of the interaction between radiation and genetic factors for sarcoma susceptibility, children whose underlying disease was retinoblastoma or a primary malignant bone tumor have an even higher risk. As an example, a report from the UK Registry of Childhood Tumors included a cohort of 13,175 pediatric patients, seen from 1940 to 1983, who had survived for at least three years following treatment of a primary tumor [13]. The estimated 20-year risk for developing bone cancer did not exceed 0.9 percent unless the underlying disease was inherited retinoblastoma (7.2 percent), Ewing sarcoma (5.4 percent), or other malignant bone tumors (2.4 percent).

Radiation-associated osteosarcomas are the most frequent second primary cancer occurring during the first 20 years following RT for a solid cancer in childhood [12]. Secondary osteosarcomas and soft tissue sarcomas are particularly common in individuals with hereditary retinoblastoma. (See 'Genetic predisposition' below.)

The mean age at diagnosis for a radiation-associated sarcoma is between 50 and 67 years for adults [26-29]. Children are typically diagnosed as teenagers or young adults, reflecting the long average latency period of between 16 and 33 years [27,30,31]. (See 'Latency period' below.)

Most common primary cancers — The most common primary cancers associated with a later radiation-associated sarcoma in adults are breast cancer, lymphoma, head and neck, and gynecologic cancers [9,11,28,29,32-36]. This distribution likely reflects the fact that larger numbers of patients are treated for, and survive, these tumors (see "Management of late complications of head and neck cancer and its treatment", section on 'Second malignancies' and "Overview of long-term complications of therapy in breast cancer survivors and patterns of relapse", section on 'Risks associated with radiation therapy' and "Treatment-related toxicity from the use of radiation therapy for gynecologic malignancies", section on 'Bone and bone marrow'). Because of the association with cancers that predominantly occur in women, they are more commonly affected, with a female to male ratio of 2:1. For gender-neutral tumors, there is no obvious female-male predisposition.

The most common primary tumors associated with a secondary radiation-associated sarcoma in children include retinoblastoma, Ewing sarcoma family of tumors, rhabdomyosarcoma, Hodgkin lymphoma, brain tumors, and Wilms tumor [12,25,37-42]. Issues related to chemotherapy exposure are addressed below. (See 'Chemotherapy agents' below.)

Histologic distribution — Postradiation sarcomas can arise in soft tissue or bone, with soft tissue sarcomas predominating. For most series that include them, bone sarcomas represent 20 to 30 percent of all radiation-associated sarcomas [9,10,26,43,44], although bone sarcomas predominate in a minority [36].

Most postradiation soft tissue sarcomas are high grade, and they can be of a variety of histologies, the most common of which are malignant peripheral nerve sheath tumor (MPNST), undifferentiated/unclassified soft tissue sarcoma (formerly a subset of tumors referred to as "malignant fibrous histiocytoma" [MFH] and now including undifferentiated pleomorphic sarcoma [UPS], as well as morphologically round cell and spindle cell variants [45]), angiosarcoma, fibrosarcoma, and leiomyosarcoma (particularly in retinoblastoma survivors) (figure 1) [9-11,26,29,31,32,46-49]. In one large series of over 69,000 five-year childhood cancer survivors, retinoblastoma survivors experienced the highest risk of developing a secondary soft tissue sarcoma (standardized incidence ratio [SIR] 72.8, 95% CI 56.1-93.0), especially leiomyosarcoma (SIR 342.9, 95% CI 245-466.9) [31]. However, 76.9 percent of the leiomyosarcomas observed after retinoblastoma treatment developed outside of the irradiated field, and they were mostly attributed to an inherited cancer predisposition syndrome. The absolute excess risk among retinoblastoma survivors increased substantially with increasing years since diagnosis, and attained age. (See 'Genetic predisposition' below.)

The most common sarcoma to occur after RT for breast cancer is angiosarcoma. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Histopathology' and "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Histologic classification'.)

Radiation-associated bone sarcomas are typically high-grade and are predominantly osteosarcomas; fibrosarcomas, undifferentiated high-grade pleomorphic sarcomas of bone (previously termed malignant fibrous histiocytomas of bone), and chondrosarcomas are diagnosed less frequently [12,50,51]. (See "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management", section on 'Histologic classification'.)

RISK FACTORS — By definition, a radiation-associated sarcoma is an iatrogenic disease that is related to prior radiation exposure. Radiation-associated sarcomas have been reported to occur as early as a few months following completion of radiation therapy (RT), to as long as 54 years later [52]. In most adult series, the average duration between RT and diagnosis of a secondary sarcoma ranges from 7 to 16 years [10,19,28,29,32,35,43,44,52-62]. Among breast cancer survivors, the average latency period for secondary postradiation breast sarcomas is 10 to 11 years, but it is variable and tends to be shorter (four to eight years) for angiosarcomas [29]. A secondary angiosarcoma has been reported as early as 14 months following breast irradiation [63]. (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Radiation exposure in breast cancer survivors'.)

Among childhood cancer survivors, the duration between RT and secondary sarcoma averages 12 to 13 years in most series, but in some, it is as low as 5.5 years [12,22,25,41,64-66].

Sarcomas after a primary diagnosis of retinoblastoma occur on average one year earlier in irradiated patients as compared with nonirradiated patients (12 versus 13 years in one series) [49]. In this series, leiomyosarcomas were the most common histology, and the majority were diagnosed 30 or more years after the retinoblastoma treatment.

A latency period is necessary to differentiate a radiation-associated sarcoma from a primary sarcoma that is unrelated to RT. However, there is no consensus as to the interval to distinguish the two. This subject is discussed in detail below. (See 'Latency period' below.)

The development of radiation-associated sarcoma may be influenced by factors such as dose, age at initial exposure, exposure to chemotherapeutic agents (particularly alkylating agents), and genetic tendency.

Radiation dose and age of exposure — The frequency of a radiation-associated sarcoma increases with the dose of radiation, as the postradiation observation period lengthens, and with younger age at the time RT was administered [22,67-70].

Sarcoma appears to be a complication of high-dose therapy, although it can be seen with lower doses (<40 Gy) [70]. To induce a tumor, the radiation dose must be great enough to cause genetic damage but not so great that it kills the cell. This often happens to tissues that are near the edge of the radiation field. Above a certain "curative" dose, cells theoretically lose their proliferative potential along with their carcinogenic capacity. However, there is great uncertainty concerning the risk for radiation-associated carcinogenesis at high doses [71]. While some animal and human data suggest a decrease in secondary sarcomas at very high doses, other data suggest a plateau in risk. A systematic review of epidemiologic studies of the radiation dose-response relationship concluded that there was little evidence that the dose-response curve was nonlinear in the direction of a downturn in risk with higher RT doses, even at organ doses ≥60 Gy [71].

It is also not known whether newer radiation techniques such as intensity-modulated RT (IMRT; which decreases the volume of normal tissue exposed to high RT doses, but increases the volume exposed to lower doses of irradiation [72,73]) or hypofractionated regimens may influence the risk of a radiation-associated sarcoma. Similarly, the use of partial breast irradiation may help limit this and other toxicities of breast irradiation. (See "Radiation therapy techniques for newly diagnosed, non-metastatic breast cancer", section on 'Accelerated partial breast irradiation' and "Radiation therapy techniques in cancer treatment", section on 'Intensity-modulated radiation therapy'.)

High-dose radiation exposure (15+ Gy)

Childhood RT — Most studies in children have shown an approximately linear correlation between RT dose and the risk of a radiation-associated sarcoma at doses above 15 Gy [12,13,15,20,22,25,74-76]. As an example, in a nested case-control study of secondary sarcomas (105 cases, 422 matched controls) derived from a cohort of 14,372 childhood cancer survivors, a dose response was observed, with elevated risks at doses between 10 and 29.9 Gy (odds ratio 15.6, 95% CI 4.5-53.9), 30 to 49.9 Gy (OR 16.0, 95% CI 3.8-67.8), and >50 Gy (OR 114.1, 95% CI 13.5-964.8) [25].

Greater risks have been associated with younger age at primary diagnosis. As an example, in a series of 108 secondary sarcomas developing in a cohort of 14,372 participants in the Childhood Cancer Survivor Study, the risk of a secondary sarcoma was more than ninefold higher amongst childhood cancer survivors when compared with the general population, and the risk was highest for patients younger than four years of age at the time of initial cancer diagnosis [77].

Exposure in adulthood — Most studies also support a relationship between RT dose and the risk of a radiation-associated sarcoma in adults [8,78-81]. As an example, in a cohort of 6597 breast cancer patients, in which 14 developed a subsequent sarcoma, compared with doses <14 Gy, the odds ratio (OR) for sarcoma in women who received 14 to 44 Gy was 1.6 (95% CI 0.2-11.0), and it was 30.6 (95% CI 4.9-611) for doses >44 Gy [78].

Risk also increases with greater time elapsed since the diagnosis of the first cancer, and in patients treated at a younger as compared with older age [27,78]. This was shown in a systematic evaluation of secondary sarcomas after any first cancer in a report derived from the United States Surveillance, Epidemiology, and End Results (SEER) registry of patients surviving at least one year after an initial non-sarcoma cancer diagnosed at age 20 to 79 between 1973 and 2008 [27]. Of the approximately 1.9 million adulthood cancer survivors, there were 1342 secondary soft tissue sarcomas and 314 secondary bone sarcomas diagnosed after an average follow-up of 13 years. The observed number of sarcomas was compared with the expected number of sarcomas in the general population, which was estimated using age and sex-specific incidence rates, and standardized incidence ratios were calculated as the observed divided by the expected number of second sarcomas for patients who received RT versus those who did not. For both types of sarcoma, the SIRs were elevated for patients who received RT versus those who did not, and the difference increased with increasing time since diagnosis of the first cancer (table 2). Patients treated at a younger age also had higher SIRs than did those treated at an older age (table 3).

Lower dose radiation exposure — In general, there is no evidence of increased sarcoma risk in patients treated with lower dose RT (<14 to 15 Gy), although the statistical power in many of these studies to detect small excess risks is generally low due to the small number of patients treated [12,26,82-85]. The Life Span Study of Japanese atomic bomb survivors provides the most informative data in this regard, with the latest data suggesting an increased risk of both secondary bone and soft tissue sarcomas with exposure to much lower doses of ionizing radiation than have previously been reported [86,87]. In the case of bone sarcomas, the mean ages at exposure and at diagnosis were 32 and 62 years, respectively, and a dose threshold was found at 0.85 Gy (95% CI 0.12-1.85 Gy) with a linear dose response association above this level [87]. For soft tissue sarcoma, the mean ages at the time of bombing and diagnosis were 27 and 64 years, respectively [86]. The mean colon dose was 0.18 Gy (95% CI 0-2.35 Gy), and the majority of sarcomas arose in the uterus and stomach. A linear dose response model was observed, but a threshold dose could not be discerned.

Potential effect modifiers

Genetic predisposition — Several rare familial genetic syndromes are associated with an increased risk of soft tissue and/or bone sarcoma, including Li-Fraumeni syndrome, retinoblastoma, Werner syndrome, neurofibromatosis type I, familial gastrointestinal stromal tumors (GIST), Costello syndrome, and Nijmegen breakage syndrome. (See "Clinical presentation, diagnosis, and prognosis of gastrointestinal stromal tumors" and "Li-Fraumeni syndrome" and "Thyroid nodules and cancer in children", section on 'Genetic predisposition' and "Neurofibromatosis type 1 (NF1): Pathogenesis, clinical features, and diagnosis" and "Rhabdomyosarcoma in childhood and adolescence: Epidemiology, pathology, and molecular pathogenesis", section on 'Inherited syndromes' and "Nijmegen breakage syndrome" and "Retinoblastoma: Treatment and outcome", section on 'Second malignancies'.)

The question of whether individuals with these heritable syndromes are more susceptible to the effects of ionizing radiation than normal individuals is of interest, but the available data to answer this question are scant. Many such patients have been excluded from clinical studies, and these syndromes affect only a small proportion of the general population. To date, however, strong evidence for enhanced radiosensitivity has been observed only for Nijmegen breakage syndrome, and mainly associated for lymphopoietic tumors [88]. (See "Nijmegen breakage syndrome".)

The situation is less clear for Li-Fraumeni syndrome and retinoblastoma. Although one cohort study of mortality from subsequent malignancy in retinoblastoma patients did not detect a significant interaction between hereditary status and treatment with RT, a large proportion of the secondary sarcomas in the irradiated patients in fact arose within the radiated field [89]. Others note the opposite; in a large cohort study of over 69,000 five-year childhood cancer survivors, 76.9 percent of the secondary leiomyosarcomas developed outside of irradiated tissue [31]. In a study of Li-Fraumeni family members, 50 percent of the secondary sarcomas occurred in the RT field [90]. (See "Li-Fraumeni syndrome", section on 'Radiation-associated cancers'.)

In particular, patients with familial (hereditary) retinoblastoma, with mutations in the (Rb) oncogene on chromosome 13q, are at particularly high risk for later development of sarcomas, not only within irradiated fields, but also in sites distant from RT ports [30,31]. With the greater use of intra-arterial chemotherapy (without radiation) as primary therapy for retinoblastoma, fewer second malignancies are expected in this setting. (See "Retinoblastoma: Treatment and outcome", section on 'Second malignancies' and "Retinoblastoma: Treatment and outcome", section on 'Local chemotherapy'.)

Chemotherapy agents — Exposure to chemotherapy, particularly alkylating agents, may potentiate the effect of previous RT in childhood cancer survivors and thus serve as another predisposing factor for the development of a post-treatment sarcoma [12,13,22,25,75,76]:

●In a series of 9719 two-year survivors of a childhood cancer from the multicenter Late Effects Study Group, each of the 64 cases who developed bone sarcoma were matched to two randomly selected controls who did not, and matched for histologic characteristics of the first tumor, duration of follow-up, age at diagnosis of the first tumor, sex, and race. [12]. After adjusting for RT, treatment with alkylating agents was linked to bone cancer (relative risk [RR] 4.7, 95% CI 1.0-22.3), with the risk rising as cumulative drug exposure increased, whether or not RT was given.

●In a series of 14,372 pediatric cancer survivors participating in the Childhood Cancer Survivor Study, in multivariate analysis, exposure to RT was associated with a 4.1-fold increased risk of developing a secondary sarcoma, and the use of chemotherapy potentiated this risk [25]. Higher total dose exposure to anthracyclines (odds ratio [OR] 3.5) or to alkylating agents (RR 2.2) continued to be associated with an increased risk of sarcoma when radiation exposure was controlled for in the regression model. However, the elevated risk seen with exposure to alkylating agents was no longer significant after adjusting for anthracycline exposure.

●In a cohort study of 4400 three-year survivors of a first solid childhood cancer diagnosed in France or the United Kingdom between 1942 and 1985, and followed on average for 15 years, there were 25 cases of soft tissue sarcoma [75]. The risk of a soft tissue sarcoma was 54-fold higher in this cohort as compared with the general population, and it was significantly higher among those treated with both chemotherapy and RT (SIR 113, 95% CI 62-185, 13 soft tissue sarcomas) compared with RT alone (SIR 19, 95% CI 3-60, 2 soft tissue sarcomas).

●At least some data suggest that treatment with anthracyclines may shorten the interval to development of a secondary bone tumor [14].

Whether chemotherapy potentiates the effects of RT in adults is less clear. One systematic review of radiation-associated sarcoma of the breast concluded that there was no evidence that prior chemotherapy was a contributing risk factor [91]. The overall lower cumulative dose of alkylating agents in breast cancer as compared with sarcoma is one hypothesis for the lack of an association in this context.

Chronic edema — Cutaneous lymphangiosarcoma is more often found in breasts developing postoperative and post-irradiation edema and fibrosis (Stewart-Treves syndrome) [92]. It mainly affects women over the age of 60 and in those who underwent axillary lymphadenectomy in addition to RT. The very rare nature of Stewart-Treves syndrome suggests that if there is an interaction between radiation and lymphedema, it is very weak. (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Lymphedema'.)

CLINICAL PRESENTATION — In general, radiation-related sarcomas present in a manner that is similar to de novo primary sarcomas of the bone or soft tissue. However, a radiation-associated sarcoma may be more difficult to identify by physical examination because of radiation therapy (RT)-associated tissue changes.

Soft tissue sarcoma — The most common presenting complaint for a soft tissue sarcoma is a gradually enlarging, painless mass in or near the radiation field. Some patients complain of pain or symptoms associated with compression by the mass, including paresthesias or edema in an extremity. The location depends on the site of prior RT; common sites include the head and neck, extremities, and retroperitoneum [35]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Clinical presentation' and "Head and neck sarcomas", section on 'Clinical presentation and diagnostic evaluation' and "Clinical presentation and diagnosis of retroperitoneal soft tissue sarcoma", section on 'Clinical symptoms'.)

Angiosarcoma is a sarcoma subtype that is strongly associated with breast cancer treatment that includes radiation of the breast or chest wall [93]. Therapy-related breast/chest wall angiosarcomas have a distinct appearance, presenting as single or, more commonly, multiple ecchymotic macular or purplish papular cutaneous lesions in the breast (picture 1) or chest wall (picture 2), or an edematous upper extremity [60,94]. Skin edema (peau d'orange) may be present. In contrast to primary angiosarcomas, which arise within the breast parenchymal tissue, secondary angiosarcomas often affect only the skin. (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Clinical features and diagnosis'.)

Angiosarcomas arising at other sites may present differently. As an example, a secondary angiosarcoma of the rectum in a patient who received RT for prostate or rectal cancer may present with rectal bleeding [95,96].

Among survivors of childhood cancer, the most common secondary soft tissue sarcoma subtypes are malignant peripheral nerve sheath tumor (MPNST), leiomyosarcoma, and fibromatous neoplasms [31].

Patients with radiation-associated soft tissue sarcoma may also present with metastatic disease. In particular, angiosarcomas metastasize early, most often to the lung and liver. They also metastasize to the lymph nodes, bones, brain, skin, and the contralateral breast, which are unusual locations for other soft tissue sarcomas, for which the most common site of metastatic spread is the lungs. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Pattern of spread' and "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Staging'.)

Bone sarcoma — Bone sarcomas present with bone pain that is often worse at night, and with swelling with or without a tender soft tissue mass adherent to bone. The most common sites of involvement are the head and neck, extremities, pelvis, and the thoracic wall, largely due to the higher frequency of first cancers in these locations (see 'Epidemiology and histologic distribution' above). As an example, in one series of 42 bone sarcomas following RT for a variety of conditions, the pelvis was the single most commonly affected site, but tumors were equally distributed between the central skeleton and limbs [51]. Nine patients presented with metastatic disease.

Following RT for breast cancer, secondary bone sarcomas are most commonly found in the scapula (40 percent), followed by the humerus (26 percent), and other thoracic bones (ribs, clavicle, sternum) [97]. Soft tissue sarcomas may occur over the anterior chest wall, parasternal area, or axilla [47]. (See "Osteosarcoma: Epidemiology, pathology, clinical presentation, and diagnosis", section on 'Clinical presentation' and "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Radiation exposure in breast cancer survivors'.)

Patients may also present with metastatic disease. As with soft tissue sarcomas, the most common site of metastatic involvement of radiation-associated bone sarcomas is the lungs. (See "Osteosarcoma: Epidemiology, pathology, clinical presentation, and diagnosis", section on 'Postdiagnostic evaluation'.)

DIAGNOSIS — Although the diagnosis may be suspected based upon imaging studies, examination of tissue is required to establish the diagnosis of a soft tissue or bone sarcoma. (See 'Biopsy' below.)

Patients with radiation-associated sarcoma often experience a delay in diagnosis because of difficulty in examination of previously irradiated tissues, lack of specific symptoms, long latency period between treatment of primary cancer and the development of the secondary sarcoma, and difficulty with obtaining an adequate biopsy [98,99].

Imaging evaluation — For patients with a soft tissue or bone mass or unexplained bone pain, plain radiographs are often the first diagnostic study to be ordered. Although findings on plain radiographs can suggest the diagnosis of malignancy by demonstrating cortical bone destruction or a mineralized soft tissue mass, the definition of tumor size and local intraosseous and extraosseous extent is most accurately achieved by cross-sectional imaging, computed tomography (CT), or magnetic resonance imaging (MRI). In a series of 63 cases of radiation-associated sarcoma, the cross-sectional imaging findings of secondary bone sarcomas included a soft tissue mass (96 percent), cortical bone destruction (83 percent), tumor matrix mineralization (48 percent), and a periosteal reaction (38 percent) [36]. For soft tissue sarcomas, radiologic findings included soft tissue mass and bone destruction; bone expansion and periosteal reaction were usually absent [36]. The presence of a soft tissue mass, particularly in combination with either bone destruction or tumour matrix mineralization is very suggestive of malignancy.

Although the definition of tumor size and local extent can be achieved using either CT or MRI, MRI is generally preferred because of its better delineation of soft tissue structures [36,100]. MRI findings of a radiation-associated sarcoma are variable and generally not diagnosis-specific [101,102].

Uncommonly, MRI findings may distinguish a radiation-associated sarcoma from a recurrence of the original malignancy [103]. Replacement of the normal marrow with fat as a consequence of radiation therapy (RT) is a particularly useful sign with MRI to suggest a RT-associated sarcoma (seen best as hyperintensity on T1-weighted imaging) [104,105]. The presence of soft tissue T1 hypointensity, rather than the expected T1 hyperintense fat, may also be an early indication of malignancy.

Imaging findings of cutaneous angiosarcoma of the breast include the development of progressive skin or trabecular thickening in an area of the breast that is separate from the patient's original breast cancer, and enhancing cutaneous nodules. Mammography may show increased skin thickening, typically without an overt mass. (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Imaging'.)

Among patients undergoing periodic radiographic surveillance after treatment of a soft tissue or bone sarcoma, radiographic change after several years of stability, new onset of symptoms or signs, and the development of osteoid or chondroid mineralization, in the absence of infection or fracture, should also raise suspicion for a secondary sarcoma. (See "Overview of multimodality treatment for primary soft tissue sarcoma of the extremities and superficial trunk", section on 'Posttreatment sarcoma surveillance' and "Chemotherapy and radiation therapy in the management of osteosarcoma", section on 'Posttreatment surveillance' and "Treatment of Ewing sarcoma".)

Biopsy — Biopsy of suspected lesions is essential for diagnosis and treatment planning. For both bone and soft tissue sarcomas, the diagnostic biopsy must be carefully planned to ensure that adequate tissue is obtained in a manner that does not compromise definitive therapy. A poorly placed initial biopsy may preclude subsequent surgical resection, preparation of flaps, and/or cosmetic repair, or result in the need for a more extensive surgery to encompass the biopsy site at the time of definitive resection. Biopsies should take place after the completion of the staging studies, and the surgeon, radiologist, and pathologist should review these studies in detail so that each member of the team is fully apprised of the diagnostic considerations. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Biopsy' and "Bone tumors: Diagnosis and biopsy techniques", section on 'Biopsy techniques'.)

Core needle biopsy is considered the preferred method to achieve an initial biopsy in most cases due to its low incidence of complications and high diagnostic accuracy obtained [106]. Although fine needle aspiration biopsy under radiologic guidance may obviate the need for open biopsy, the risk of a nondiagnostic or nonrepresentative sample must be considered. (See "Bone tumors: Diagnosis and biopsy techniques", section on 'Planning the biopsy'.)

If an open biopsy is needed, the incision should be placed in accordance with the planned surgical resection; the primary tumor and the entire biopsy tract should be resected en bloc. Meticulous hemostasis and the judicious use of a drain are important to avoid the spread of hematoma-containing tumor cells. If a soft tissue mass is not present, or material is nondiagnostic, a bone defect may be required to obtain tissue. If so, it should be a small, oval defect, and a polymethylmethacrylate plug may be used to close the hole in order to minimize hematoma. (See "Bone tumors: Diagnosis and biopsy techniques", section on 'Operative biopsy'.)

There are no specific histopathologic criteria for determining whether a sarcoma that arises in an irradiated field is a primary or secondary sarcoma. This subject is discussed in detail below. (See 'Radiation-associated versus sporadic sarcoma' below.)

DIFFERENTIAL DIAGNOSIS

Sporadic sarcoma — The main differential is with a sporadic sarcoma that may be primary or secondary. Not all sarcomas that arise in an irradiated field are radiation-associated sarcomas [107]. Sarcomas as secondary tumors can arise without a radiation history, as an effect of other treatments, or as a treatment complication such as chronic edema or genetic predisposition. As examples:

●Angiosarcoma is a vascular sarcoma that can arise in the setting of chronic lymphedema, usually as a consequence of prior treatment for a malignancy. Angiosarcomas of the upper extremity, breast, and axilla arising in women with chronic lymphedema after breast cancer therapy were initially described by Stewart and Treves using the term lymphangiosarcoma, and this syndrome is now designated Stewart-Treves syndrome [108]. Although such women typically present with longstanding extensive arm edema after mastectomy and axillary lymph node dissection, it may occur after radiation therapy (RT) alone. In these cases, it results from axillary lymph node sclerosis, and as such, angiosarcoma represents a secondary sarcoma, but it does not meet the criteria for a radiation-associated sarcoma. (See 'Radiation-associated versus sporadic sarcoma' below and "Clinical staging and conservative management of peripheral lymphedema", section on 'Lymphangiosarcoma'.)

●Sarcomas are associated with family cancer syndromes and may arise as secondary tumors without a radiation history. As an example, secondary osteosarcomas are relatively common following RT for hereditary retinoblastoma in childhood, but many arise within unirradiated tissues. (See 'Genetic predisposition' above.)

Radiation-associated versus sporadic sarcoma — There are no specific histopathologic criteria for determining whether a sarcoma that arises in an irradiated field is a primary or secondary sarcoma, although the morphology of adjacent tissues may be suggestive if it shows radiation-related changes (eg, dense cellular fibrosis, atypical fibroblasts, alteration of the vascular architecture, and abundant fibrous stroma in the dermis adjacent to the sarcoma) [107]. As with all pathologic analyses, histology must be interpreted in the context of a clinical history, which is often not provided to pathologists.

The presumptive diagnosis of a sarcoma as a secondary, treatment-related tumor rather than a sporadic sarcoma is based upon criteria initially proposed in 1948 by Cahan [109] and later revised in 1971 and 1999 [53,110]:

●Radiation must have been given previously, with the sarcoma arising in the area included within the 5 percent isodose line of the radiated field.

●The histologic features of the original lesion and the postradiation sarcoma should be different.

●Patients with inherited syndromes that predispose to sarcomas even in the absence of RT, such as Li-Fraumeni or Rothmund-Thomson syndrome, should be excluded.

●Initially, the minimal latency period (the time period in years between the initiation of RT and the histologic diagnosis of the second neoplasm) was suggested to be over five years [109], but subsequently, shorter time frames have been thought to be adequate. (See 'Latency period' below.)

We also note that it is difficult to consider a sarcoma that harbors a chromosomal translocation as being radiation-associated, given the specific nature of that change in comparison with the aneuploid karyotype seen for the vast majority of radiation-associated sarcomas. (See 'Radiation signatures' below and "Pathogenetic factors in soft tissue and bone sarcomas", section on 'Chromosomal translocations'.)

Latency period — A latency period is necessary to differentiate an RT-associated sarcoma from a sporadic sarcoma that post-dated RT, although the best interval to establish this distinction is debated:

●Cahan established that a five-year latency period would distinguish between a radiation-related and a sporadic sarcoma, but this was based upon few case reports, presumed cure of the primary tumor, and was intentionally stringent [109].

●Arlen, et al published a modification that used at least three to four years latency, also based upon case reports [53], and was subsequently used to define a radiation-associated sarcoma by several series [26,54].

●Laskin, et al further modified this to at least two years latency in a study of 53 patients with radiation-associated soft tissue sarcoma [35].

●In a study from Cha, et al, a cutoff of at least six months was chosen, although the median interval between radiation exposure and sarcoma formation in their series was 8.4 years [32], a long median interval between RT exposure and secondary sarcoma has been observed by others [16].

We (and others [62]) believe that an equally arbitrary latency period of three years is reasonable to define radiation association, except in the case of angiosarcoma following breast cancer treatment, recognizing that a latency period as short as six months has been observed [111]. (See "Breast sarcoma: Epidemiology, risk factors, clinical presentation, diagnosis, and staging", section on 'Radiation exposure in breast cancer survivors'.)

Radiation signatures — There has been considerable interest in identifying a molecular/genetic expression profile that can differentiate between radiation-related and sporadic sarcomas. Most of the studies in this field have used some modification of the 1948 Cahan criteria for classifying sarcomas as radiation related [109]. While satisfying these criteria is likely to result in a high probability that the sarcoma is radiation related, there remains no gold standard for defining a radiation-associated sarcoma. (See 'Radiation-associated versus sporadic sarcoma' above.)

●Early work used conventional cytogenetic analysis (eg, comparative genomic hybridization) to identify large-scale chromosomal abnormalities, but only loss of material on chromosomes X and 13 was observed; most radiation-associated sarcomas displayed relatively complex karyotypes with multiple, mostly unbalanced, structural rearrangements [112,113].

●More studies using polymerase chain reaction followed by direct sequencing found a high rate of TP53 mutations in radiation-associated compared with sporadic sarcomas (88 versus 20 percent, and 58 versus 17 percent for two different series, respectively [114,115]). In a later study from one of these groups, 12 of 36 cases were secondary to retinoblastoma, a syndrome caused by mutations in the RB1 gene [116]. The authors hypothesized that the inactivation of the p53 allele in 12 of 36 cases was due to irradiation rather than loss of RB1 expression. They also noted that in 40 percent of the radiation-associated tumors, neither pathway was inactivated, suggesting the presence of further unidentified pathways involved in radiation carcinogenesis.

●Cutaneous angiosarcomas after treatment for breast cancer often show amplification of MYC at locus 8q24.21, a finding that has also been seen in lymphedema-associated angiosarcomas, but not primary angiosarcomas [117-119].

●Using a microarray analysis approach, others were able to generate a molecular signature of 135 genes and were able to discriminate radiation-associated sarcomas from sporadic sarcomas with 96 percent sensitivity and 62 percent specificity [120]. Examination of the specific genes suggested a particular role for mitochondrial genes and genes involved in detoxification or antioxidant pathways, suggesting that mitochondrial dysfunction of chronic oxidative stress could be hallmarks of radiation-associated tumors.

While promising, these approaches are not yet ready for clinical use.

Other diagnoses — In addition to sporadic sarcoma, the differential diagnosis of a mass developing in a previously irradiated field includes inflammatory lesions (eg, arthritis, bursitis, gout), other malignancy (recurrence of the original tumor, metastases, a radiation-associated carcinoma, new lymphoma or melanoma), and radiation necrosis of bone (osteonecrosis). These diagnoses may be made based upon imaging findings, but a biopsy is usually required to distinguish between malignancy types.

The source of musculoskeletal pain in patients with cancer is often elusive. The differential diagnosis may include arthritis, bursitis/tendinitis, gout and pseudogout, metastases from original tumor, or metastases from a new second tumor. Pain within a field of radiation could also be due to a stress fracture or radiation osteitis. These diagnoses may be made based upon imaging findings (eg, in radiation osteitis, there is not usually an associated soft tissue mass [36]), but a biopsy is usually required to distinguish between malignancy types.

STAGING

Staging system — Radiation-associated soft tissue sarcomas are staged like other soft tissue sarcomas using the tumor, node, metastasis (TNM) criteria and stage groupings of the joint American Joint Committee on Cancer (AJCC) and Union for International Cancer Control (UICC). The most recent version (2017, eighth edition) contains separate T criteria for sarcomas arising in the head and neck (table 4), extremities/trunk (table 5), abdominal/thoracic viscera (table 6), and retroperitoneum (table 7) [121]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Staging'.)

Staging and prognostication stratification schema for rhabdomyosarcoma are discussed in detail separately. (See "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging", section on 'Staging and prognostic stratification'.)

Several staging systems have been used for bone sarcomas (see "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management", section on 'Tumor staging'):

●There is an AJCC/UICC TNM staging system for bone sarcomas, but it has not been widely utilized in the past. The most recent version (eighth edition, 2017) has TNM criteria and prognostic stage groupings for bone sarcomas arising in the appendicular skeleton, trunk, skull, and facial bones; tumors arising in the pelvis and spine have separate and distinct TNM classifications, but no specific stage groupings (table 8) [122]. Whether this version will be widely adopted for clinical use remains to be seen.

●Another staging system, used by some orthopedic oncologists for de novo and secondary bone sarcomas, was developed at the University of Florida and based upon a retrospective review of cases of primary malignant tumors of bone treated by primary surgical resection (table 9) [123,124]. This system characterizes nonmetastatic malignant bone tumors by grade (low grade [stage I] versus high grade [stage II]), and further subdivides these stages according to the local anatomic extent, which is determined by whether the tumor extends through the cortex of the involved bone. (See "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management", section on 'Tumor staging'.)

●Ewing sarcoma as a radiation-associated sarcoma is an extremely rare diagnosis, given the nature of the genetic change that gives rise to Ewing sarcoma as compared with most radiation-associated sarcomas. The staging classification used for Ewing sarcoma is complex and discussed in detail separately. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Staging system'.)

Staging workup — The lung is the predominant site of metastases for both soft tissue and bone tumors. Computed tomography (CT) of the chest is typically performed to detect pulmonary metastases in patients with either a soft tissue or bone sarcoma. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Evaluation for metastatic disease' and "Osteosarcoma: Epidemiology, pathology, clinical presentation, and diagnosis", section on 'Evaluation for systemic disease'.)

For bone sarcomas, guidelines from the National Comprehensive Cancer Network (NCCN) [125] suggest either an FDG-PET scan and/or bone scan in the workup of a suspected osteosarcoma, and imaging guidelines from the Children's Oncology Group Bone Tumor Committee for both osteosarcoma and Ewing sarcoma recommend radionuclide bone scan and/or PET scan for whole body staging. This subject is addressed in detail separately. (See "Osteosarcoma: Epidemiology, pathology, clinical presentation, and diagnosis", section on 'Evaluation for systemic disease' and "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Metastatic work-up'.)

PROGNOSIS AND TREATMENT — The prognosis of patients with secondary radiation-associated sarcomas is poorer than that of patients with primary sarcomas, with most series reporting overall five-year survival rates in the range of 10 to 50 percent [16,46,126]. As examples:

●In a comparison of 130 patients with radiation-associated soft tissue sarcomas with a matched cohort of sporadic sarcomas, patients with a secondary sarcoma had significantly worse disease specific survival (hazard ratio [HR] 1.7, range 1.1-2.4) [46]. For one specific histologic subtype (malignant fibrous histiocytoma [MFH] now termed undifferentiated pleomorphic sarcoma [UPS]), five-year disease specific survival was 44 percent for patients with radiation-associated sarcomas compared with 66 percent for patients with sporadic tumors.

●Similarly, in another case-control study comparing sarcoma-related survival of 98 patients with radiation-associated sarcoma with that of 239 sporadic high-grade malignant sarcomas, the five-year survival was significantly worse for the secondary tumors (32 versus 51 percent, p<0.001) [126]. Among the significant differences between the groups that may have explained the worse outcomes were central location (60 percent of the radiation associated sarcomas versus 23 percent of the sporadic tumors), achievement of complete surgical resection (46 versus 69 percent), and radiation therapy (RT) used less often (22 versus 33 percent). In addition to inferior survival, local recurrence rates were also higher in the radiation-associated sarcomas (41 versus 17 percent).

Nonetheless, patients with radiation associated sarcomas warrant aggressive treatment, as secondary sarcomas are still potentially curable [26,29,32,43,51,54,65,127,128].

There are several potential reasons for the poor outcomes, including delay in diagnosis, the preponderance of central versus peripheral tumors, higher rates of local recurrence due to limitations on resectability due to size and location of the lesion, with an inability to achieve wide surgical margins, the high grade nature of most radiation-associated sarcomas, inability to give full-dose RT to a site that has been previously irradiated, and use of prior chemotherapy for the first cancer may limit choices for adjuvant therapy or metastatic disease [1,16,17,26,32,46,47,51,126].

In most (but not all [9]) series, extremity tumors have a better prognosis than central tumors (ie, those involving the vertebral column, pelvis, and shoulder girdle) [32,51,54,129]. In a review of 78 patients from the Mayo clinic, the five-year survival rates for radiation-associated extremity versus central tumors were 30 versus 4 percent [129].

Surgical resection is a prerequisite for cure. In a report of 80 patients with histologically verified sarcomas within a previously irradiated field, overall survival rates at two and five years were 69 and 39 percent for patients undergoing surgery, versus 10 and 0 percent for those treated with chemotherapy alone [43]. Surgical resection with wide margins offers the best opportunity for cure, although this is often not possible [29,32,129,130]. Diagnosis and management at centers of excellence in the multidisciplinary care of sarcomas may permit the best possible outcomes.

Local treatment — The main treatment for post-irradiation soft tissue sarcomas is wide local resection. If wide resection is not feasible, other options include marginal limb sparing surgery plus reirradiation, or amputation.

For patients with potentially resectable radiation-associated soft tissue and bone sarcomas, outcomes are best if a wide surgical margin can be achieved:

●In a series of 123 patients treated at Memorial Sloan Kettering Cancer Center over a 20-year period who met the criteria for a radiation-associated sarcoma, the most common malignancy for which radiation was used was breast cancer (29 percent), followed by lymphoma (16 percent) and prostate cancer (15 percent) [32]. One hundred fourteen patients were resected with curative intent by wide excision, limb sparing surgery, or amputation; 46 percent had a grossly or microscopically positive resection margin. Reirradiation was not administered. Among the resected patients, the five-year disease-specific survival rate was 41 percent, and the median disease-specific survival was 48 months (figure 2). On multivariate analysis, tumor grade and grossly positive resection margin were predictive of poor sarcoma-specific survival (figure 3 and figure 4).

●A Mayo Clinic analysis including 130 patients evaluated for a postradiation sarcoma over a 60-year period (1933 to 1992) [54]. The histologic features consisted of 83 osteosarcomas (62 percent), 32 fibrosarcomas (24 percent), 13 undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma (now referred to as undifferentiated/unclassified soft tissue sarcoma, 10 percent) and five chondrosarcomas (4 percent). Of the 61 patients with potentially resectable disease, 49 had amputations, and 12 had limb salvage procedures; the amputation rate was much higher than what is typically seen for primary soft tissue sarcomas (approximately 5 percent). External beam irradiation was delivered to 30 patients early in the series, and chemotherapy was given to a total of 55 patients (42 percent). The five-year cumulative survival rate was 68 percent for patients with peripheral (extremities, including proximal femur and hip) resectable lesions and 27 percent for patients with central (pelvis, head and neck, and ribs) resectable lesions. The local recurrence rate correlated with the surgical margin achieved; it was highest for intralesional and marginal excisions (73 and 64 percent, respectively), and lowest for patients undergoing wide local excision (23 percent).

●In a series of 42 patients treated at the Royal Orthopaedic Hospital (71 percent osteosarcoma), treatment was by surgery and chemotherapy when indicated; 30 patients (71 percent) were treated with the intention to cure [51]. The survival rate was 41 percent at five years for those treated with the intention to cure, but in those treated palliatively, the mean survival was only 8.8 months (2 to 22), and all had died by two years. The only factor found to be significant for survival was the ability to completely resect the tumor.

●Among a series of 176 patients with localized radiation-associated sarcoma treated at Brigham and Women's Hospital and Dana-Farber Cancer Institute, 91 percent underwent resection. Resection margin status was available for 147 of the 161 patients who underwent surgery. Three-year overall survival was higher for patients with negative margins compared to those with positive margins (90 versus 66 percent) [29].

Despite aggressive attempts at curative surgery, it is not surprising that both local and distant relapse rates are high. In a retrospective review of 67 patients diagnosed with a radiation-associated soft tissue sarcoma at Royal Marsden Hospital between 1990 and 2005, 56 percent were classified as high grade, 31 percent intermediate grade, and 13 percent low grade [62]. The most common histology was leiomyosarcoma. Of 67 patients, surgery was the primary treatment modality in 48, but only in 34 was the intent potentially curative. Of the 34 patients, microscopically clear margins were achieved in 75 percent of cases. No patient received adjuvant RT, but seven received adjuvant/neoadjuvant chemotherapy. At a median follow-up of 53 months, among the patients undergoing surgery with curative intent, the median sarcoma-specific survival was 54 months (two- and five-year survival: 75 and 45 percent, respectively); however, the local relapse rate was high (65 percent) as was the distant recurrence rate (44 percent). As a result, the two- and five-year recurrence-free survival rates were 41 and 25 percent, respectively.

Among patients who develop an angiosarcoma after breast-conserving therapy, radical mastectomy offers the best chance for cure [29,131]. Recurrence rates can be high, particularly for those who present with multiple skin lesions rather than a single lesion [58,91,94,132]. However, studies have shown that extensive resection with removal of all (or as much as possible of) previously irradiated skin yields favorable results [29,132,133]. In one observational series, the three-year overall survival for patients with radiation-associated angiosarcoma of the breast was 84 percent [29]. Issues related to resection for angiosarcomas following breast cancer treatment are addressed in detail separately. (See "Breast sarcoma: Treatment", section on 'Surgery'.)

Wide local resection may be particularly challenging for radiation-related sarcomas arising in the head and neck. (See "Head and neck sarcomas", section on 'Surgical principles'.)

Reirradiation — There are very few literature reports of reirradiation for sarcoma arising in a previously irradiated field [28,131,134-140]. In general, the benefits of reirradiation (which in many series are limited to lower rates of local recurrence and no clear impact on survival [28]) must be balanced against the risks of retreatment.

There are only limited reports of successful re-irradiation of patients with radiation-associated sarcomas, generally in conjunction with surgery [141]. Radiation for these patients has also been combined hyperthermia, generally in patients with chest wall sarcomas amenable to hyperthermia [142]. The risks of reirradiation to normal tissue are considerably higher than in patients who have not previously been irradiated, and, hence, great care is required to limit the dose and volume to normal tissues at risk. Some of the general principles in patients undergoing reirradiation include the use of hyperfractionated RT with smaller daily doses [142], protons, including intensity-modulated protons [143,144], brachytherapy [145], use of previously unirradiated normal tissue flaps for the surgical resections [145], and the use of chemotherapy in association with lower-dose RT [146].

In a combined literature review of 35 patients who underwent reirradiation either in the neoadjuvant or salvage setting with or without hyperthermia, a response to treatment was reported in 12 of 23 patients in whom it could be assessed (52 percent) [142]. Of the 21 patients who underwent reirradiation with or without hyperthermia for macroscopic disease, the local relapse-free survival was 29 percent (6 of 21), and four of these patients remained alive at periods from 11 to 39 months. In 14 patients, reirradiation with or without hyperthermia was used in conjunction with radical surgery as neoadjuvant or adjuvant therapy; and 7 of the 14 remained without local recurrence until death or last follow-up for a median of 38 months. Treatment-related toxicity was not reported.

Reirradiation using a hyperfractionated schedule of treatment administration has been used in the neoadjuvant or adjuvant setting in some patients with favorable results; most of the data are in patients with a radiation-related breast sarcoma following treatment for breast cancer [134,135]. There are no randomized studies to suggest this form of radiation as superior to daily fractions of radiation. These are challenging patients, and treatment is always individualized.

Role of chemotherapy

Bone sarcomas — Radiation-associated primary bone sarcomas are generally treated with chemotherapy and surgery when feasible, as are primary tumors of bone. (See "Chemotherapy and radiation therapy in the management of osteosarcoma", section on 'Overview of primary management'.)

Because of the rarity of radiation-associated osteosarcoma, very few articles report the outcome on this particular entity. There are no randomized trials addressing the benefit of adjuvant or neoadjuvant chemotherapy. Some authors suggest that chemotherapy is not helpful, as it is in primary osteosarcoma, whereas others report that, provided local control is achieved, radiation-associated osteosarcomas treated with chemotherapy and surgery have a similar prognosis to primary osteosarcoma:

●In the Rizzoli Institute experience with 20 cases of radiation-associated osteosarcoma, patients received one preoperative cycle of cisplatin, doxorubicin, and methotrexate and a second cycle with cisplatin, ifosfamide and methotrexate because of the total prior dose of doxorubicin for the primary malignancy; three postoperative cycles of cisplatin, ifosfamide, and methotrexate were given [128]. At a mean follow-up of 11 years (range 7 to 22 years), nine patients remained disease free, 10 died from osteosarcoma, and one died from a third neoplasm (acute myeloid leukemia). Event-free survival was not significantly different from that achieved in 754 patients with conventional sporadic osteosarcoma who were treated in the same period with conventional treatment protocols, although overall survival was numerically worse (five-year event-free survival: 40 versus 60 percent, p = not statistically significant; five-year overall survival: 40 versus 67 percent, p<0.01). This sporadic osteosarcoma group had an 18 percent three-year event-free survival after treatment of relapse versus 0 percent in the radiation-associated osteosarcoma group.

●In a retrospective study of 23 children who developed an osteosarcoma within the radiation treatment volume, there were 21 conventional osteosarcoma cases, one case of high-grade surface osteosarcoma, and one of periosteal osteosarcoma [147]. The sites of involvement were the craniofacial bones in six cases, the first cervical vertebra in one, the pelvic girdle in seven, and the extremities of long bones in nine. Three patients had metastatic disease at diagnosis. The aim of treatment was curative for 16 patients, and 14 had intensive chemotherapy before or after surgery. Fifteen patients achieved a complete remission. Overall and event-free survivals at eight years were 50 and 41 percent, respectively. In the subgroup of patients without metastases at diagnosis, the respective survival rates (57 and 48 percent, respectively) were slightly lower than those reported in series of patients with de novo osteosarcoma. Complete surgical resection was an important prognostic factor: None of the eight patients treated with chemotherapy alone or chemotherapy and incomplete tumor resection survived, compared with 12 survivors of the 15 patients who underwent complete resection.

●Secondary osteosarcomas of the calvarium and skull base can be particularly challenging. In a series of 16 such patients who were treated at MD Anderson Cancer Center over a 32-year period, the average age at diagnosis was 35 years old, and the median latency period was 12.5 years; the primary tumor was a brain tumor, leukemia, retinoblastoma, rhabdomyosarcoma, or nasal cavity squamous cell cancer [57]. Of the 14 who underwent resection, nine were gross-total (five with histologically confirmed negative margins), and five underwent subtotal resection; all patients received adjuvant chemotherapy. Despite aggressive therapy, only one of the 14 remained without evidence of recurrence after 51 months of follow-up.

Soft tissue sarcoma — Systemic chemotherapy is a routine component of treatment for several soft tissue sarcomas that occur predominantly in children (ie, rhabdomyosarcoma, Ewing sarcoma, including those that arise in previously irradiated sites [65]). These are all very uncommon as second malignancies. (See "Rhabdomyosarcoma in childhood, adolescence, and adulthood: Treatment", section on 'Chemotherapy' and "Treatment of Ewing sarcoma".)

In contrast, the role of adjuvant chemotherapy for the more common adult subtypes of soft tissue sarcoma (such as liposarcoma, synovial sarcoma, leiomyosarcoma, and angiosarcoma) is uncertain. (See "Adjuvant and neoadjuvant chemotherapy for soft tissue sarcoma of the extremities".)

Reiterating data from the treatment of primary disease:

●An analysis from the Sarcoma Meta-Analysis Collaboration (SMAC) suggests a significant 11 percent improvement in survival for doxorubicin and ifosfamide-based adjuvant chemotherapy compared with resection alone [148].

●However, a pooled analysis of individual patient data from the two largest adjuvant trials of doxorubicin and ifosfamide-based chemotherapy (both performed by the European Organisation for Research and Treatment of Cancer (EORTC), only one of which was included in the SMAC meta-analysis) was negative for any survival advantage.

●There is no evidence that adjuvant chemotherapy is more beneficial for the chemotherapy-sensitive subtypes of soft tissue sarcoma.

The above studies focus on primary sarcoma, not radiation-associated sarcoma, in which people have often received prior chemotherapy as well. There are few primary data addressing the benefit of chemotherapy in radiation-associated sarcomas. The contribution of chemotherapy to outcomes was addressed in a retrospective analysis of 80 cases of radiation-associated sarcomas diagnosed and treated from 1975 to 1995; 70 percent were soft tissue sarcomas, and the remainder bone [43]. Treatment included surgery alone (n = 28), surgery plus chemotherapy (n = 18), surgery plus RT with or without chemotherapy (n = 13), chemotherapy alone (n = 15), RT alone (n = 1), and no therapy (n = 5). Median survival in the entire cohort was 23 months, and two- and five-year recurrence-free survival rates were 54 and 32 percent. Overall survival was shortest in those patients undergoing chemotherapy alone (median six months), and longest for those who underwent surgery alone (median 42 months). It was intermediate in the patients who underwent surgery plus chemotherapy (median 28 months), but the authors noted a wide variety of different chemotherapy drugs, which were not always given in an optimal manner, and limited by performance status.

Angiosarcoma — Angiosarcomas unfortunately have a particularly high rate of distant recurrence, including sites such as the lung, liver, bone marrow, and brain. In one report of 95 cases of radiation-associated angiosarcoma following surgery and RT for breast cancer, after a median follow-up of 10.8 years, local recurrences developed in 48 percent of patients, while distant metastases occurred in 27 percent [60]. There are no trials specifically addressing the benefit of adjuvant chemotherapy for breast sarcomas, including radiation-associated angiosarcomas. Results from retrospective series are mixed, and as a result, it is difficult to conclude from any of these series that adjuvant chemotherapy is definitely beneficial.

Nevertheless, at some institutions, patients with high-risk sarcomas, such as large, high-grade secondary angiosarcomas, have been offered adjuvant chemotherapy if they have a good functional status. However, there are no data to support this treatment strategy. (See "Breast sarcoma: Treatment", section on 'Adjuvant chemotherapy'.)

Treatment of metastatic disease — Treatment of patients with metastatic disease follows guidelines developed for other metastatic sarcomas, with the understanding that there may be limitations on the systemic agents that can be offered on the basis of prior drug exposure or other toxicities from prior treatment. Limited metastatic disease to lung can be resected with curative intent, given the small proportion of patients with osteosarcoma or soft tissue sarcomas five years after resection of pulmonary metastatic disease. (See "Overview of the initial treatment of metastatic soft tissue sarcoma" and "Surgical resection of pulmonary metastases: Benefits, indications, preoperative evaluation, and techniques" and "Surgical resection of pulmonary metastases: Outcomes by histology".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)

SUMMARY AND RECOMMENDATIONS

●General principles of radiation-associated sarcomas – Prior exposure to ionizing radiation is a known risk factor for secondary sarcomas involving both soft tissue and bone. Although the use of radiation therapy (RT) increases the risk of sarcoma compared with no irradiation, the absolute risk is low overall (1 percent or less). Children are more susceptible to radiation-associated sarcoma than are adults. (See 'Epidemiology and histologic distribution' above.)

●Common histologies – Most postradiation soft tissue sarcomas are high grade, and they can be of a variety of histologies, the most common of which are undifferentiated/unclassified soft tissue sarcoma (formerly called malignant fibrous histiocytoma [MFH] or undifferentiated pleomorphic sarcoma [UPS]), angiosarcoma, fibrosarcoma, and leiomyosarcoma. The most common sarcoma to occur after RT for breast cancer is angiosarcoma. In series that report them, radiation-associated bone sarcomas account for approximately 20 to 30 percent of radiation-associated sarcomas, and the most common histology is osteosarcoma. (See 'Histologic distribution' above.)

●Risk factors – Radiation-associated sarcomas have been reported to occur as early as a few months following completion of RT, to as long as 54 years. The average latency period in adults ranges from 7 to 16 years; it is shorter for children (5.5 to 12 years) and for angiosarcomas after breast cancer treatment (four to eight years). (See 'Risk factors' above.)

The development of radiation-associated sarcoma may be influenced by factors such as dose, age at initial exposure, exposure to chemotherapeutic agents (especially alkylating agents), and genetic tendency. (See 'Radiation dose and age of exposure' above and 'Potential effect modifiers' above.)

●Clinical presentation – In general, radiation-related sarcomas present in a manner that is similar to sporadic primary sarcomas of the bone or soft tissue. However, the diagnosis is often delayed, in part because radiation-associated sarcomas may be more difficult to identify by physical examination because of RT-associated tissue changes. (See 'Clinical presentation' above.)

●Diagnostic evaluation

•Imaging – Although findings on plain radiographs can suggest the diagnosis of malignancy by demonstrating cortical bone destruction or a mineralized soft tissue mass, the definition of tumor size and local intraosseous and extraosseous extent is most accurately achieved by cross-sectional imaging, computed tomography (CT), or magnetic resonance imaging (MRI). (See 'Imaging evaluation' above.)

•Biopsy – Examination of tissue is required to establish the diagnosis of a soft tissue or bone sarcoma. The diagnostic biopsy must be carefully planned to ensure that adequate tissue is obtained in a manner that does not compromise definitive therapy. Core needle biopsy is considered the preferred method to achieve an initial biopsy in most cases. (See 'Biopsy' above.)

●Radiation-associated and sporadic sarcoma – There are no specific molecular or pathologic markers to guide distinction between radiation-associated and sporadic sarcomas in the radiation field. The presumptive diagnosis of a sarcoma as radiation associated and not sporadic is made on the basis of the following criteria (see 'Radiation-associated versus sporadic sarcoma' above):

•RT must have been given previously, with the sarcoma arising in the area included within the 5 percent isodose line of the radiated field.

•The histologic features of the original lesion and the postradiation sarcoma should be different.

•Patients with inherited syndromes that predispose to sarcomas even in the absence of RT, such as Li-Fraumeni or Rothmund-Thomson syndrome, should be excluded.

•Initially, the minimal latency period (the time period in years between the initiation of RT and the histologic diagnosis of the second neoplasm) was suggested to be over five years, but subsequently, shorter time frames (at least three years) have been thought to be adequate. (See 'Latency period' above.)

●Staging work-up – The lung is the predominant site of metastases for both soft tissue and bone sarcomas, whether primary or radiation-associated. CT of the chest is typically performed to detect pulmonary metastases in patients with either a soft tissue or bone sarcoma. Positron emission tomography (PET) or bone scan is recommended for patients with bone sarcoma. (See 'Staging workup' above.)

●Prognosis and treatment – The prognosis of patients with radiation-associated sarcomas is poorer than that of patients with primary sarcomas. However, aggressive management is warranted because secondary sarcomas are potentially curable. (See 'Prognosis and treatment' above.)

•Surgical resection – The main treatment is surgery with wide local excision. Outcomes are best if a wide surgical margin can be achieved. (See 'Local treatment' above.)

•Reirradiation – There are very few literature reports of reirradiation for sarcoma arising in a previously irradiated field. In general, the benefits of reirradiation must be balanced against the risks of retreatment. (See 'Reirradiation' above.)

•Chemotherapy – Radiation-associated bone sarcomas are generally treated with chemotherapy in addition to surgery, as are primary tumors of bone. In addition, systemic chemotherapy is a routine component of treatment for several soft tissue sarcomas that occur predominantly in children (ie, rhabdomyosarcoma, Ewing sarcoma), although these soft tissue sarcoma subtypes are particularly rare as radiation-associated sarcomas. (See 'Role of chemotherapy' above.)

As is the case with primary soft tissue sarcomas, the role of adjuvant chemotherapy for the more common adult subtypes of radiation-associated soft tissue sarcoma (such as UPS, synovial sarcoma, leiomyosarcoma, angiosarcoma) remains uncertain, and this approach cannot be adopted as a standard treatment, especially since patients may have received prior chemotherapy. Decisions about chemotherapy must be individualized. (See "Adjuvant and neoadjuvant chemotherapy for soft tissue sarcoma of the extremities".)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Thomas F DeLaney, MD, who contributed to an earlier version of this topic review.