Treatment of Ewing sarcoma

INTRODUCTION — 

Ewing sarcoma (ES) is a rare primary bone malignancy and most often develops in the long bones of the extremities or the flat bones of the pelvis. Less commonly, it arises in soft tissue (extraosseous ES). Multimodality therapy is the standard of care for most patients as this approach has markedly improved long-term survival. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Prognostic factors'.)

This topic will discuss the management of ES. The pathology, molecular genetics, clinical presentation, diagnosis, and prognosis are discussed separately.

●(See "Epidemiology, pathology, and molecular genetics of Ewing sarcoma".)

●(See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma".)

PRETREATMENT EVALUATION — 

The initial management of patients depends on an accurate evaluation and staging to determine if the disease is localized or there are any sites of metastases. This is discussed in more detail separately. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Staging evaluation'.)

DEFINITION OF LOCALIZED DISEASE — 

Localized ES refers to disease confined to the site of origin and without evidence of distant metastatic disease.

Metastatic disease refers to disease anywhere other than the primary site and can include involvement of lung, bone, bone marrow, distant lymph nodes, lesions discontinuous from the primary tumor (ie, skip metastases), or other distant sites.

Patients with regional lymph node involvement have generally been included in trials for patients with localized ES.

MULTIMODALITY THERAPY FOR LOCALIZED DISEASE — 

Patients with localized ES are treated with multimodality therapy (ie, chemotherapy plus surgery and/or radiation therapy (algorithm 1)).

Although <25 percent of patients have overt metastases at the time of diagnosis, ES is considered a systemic disease. Patients undergoing local therapy alone have a high relapse rate (80 to 90 percent) suggesting that most patients have subclinical metastatic disease at the time of diagnosis, even in the absence of overt metastases. Chemotherapy can successfully eradicate these deposits, and modern treatment plans all include chemotherapy, usually administered prior to (neoadjuvant) and following (adjuvant) local treatment.

For patients with localized disease, the addition of intensive multiagent chemotherapy to local therapy has had a dramatic impact on survival, and reported five-year survival rates are approximately 70 percent [1-10].

While they constitute a minority of patients, adults and those with extraosseous ES are treated in a similar fashion to other patients with ES except as specifically described below:

●Adult patients – Treatment of adults with ES should generally be guided by the same general principles as those used for younger individuals, with therapy modification based upon comorbidities and treatment tolerance. However, few clinical trials address treatment in adults since most studies have excluded older individuals. (See 'Patients 18 years or older' below.)

Less than 5 percent of cases of ES arise in adults over the age of 40. Some studies have defined "older" patients as those over the age of 15 at initial diagnosis, while others have used a cutoff of 18 years. Although some studies suggest that older age is an adverse prognostic factor in ES, it is not clear if the worse prognosis in older individuals is due to biologic differences or differences in treatment regimens (ie, regimens used in pediatric versus those used in adult medical oncology). (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Age'.)

●Extraosseous Ewing sarcoma – Patients with extraosseous ES should be treated with the same protocols used for osseous ES as a similar response to multimodality therapy has been demonstrated [7,11-14].

Neoadjuvant chemotherapy — Neoadjuvant chemotherapy is a standard part of ES treatment, given to reduce local tumor volume, facilitate resection, and treat micrometastatic disease. Reduction in tumor volume may enable limb-sparing procedures for extremity lesions, but it may also be important for rib, chest wall, and vertebral primaries [15-17].

Clinical evidence that chemotherapy is effectively treating the tumor include relief of tumor-related pain, decrease in tumor size, fall in lactate dehydrogenase level, radiologic improvement, and evidence of necrosis in the resected specimen.

Patients <18 years

VDC/IE preferred — For patients under 18 years with localized ES, we recommend interval-compressed neoadjuvant chemotherapy with alternating cycles of vincristine/doxorubicin/cyclophosphamide and ifosfamide/etoposide (VDC/IE) given every 14 days with hematopoietic growth factor support (table 1). Patients are treated for a total of six cycles (12 weeks) prior to local therapy. VDC/IE is preferred over vincristine, ifosfamide, doxorubicin, and etoposide (VIDE) because it improves overall survival (OS), has a shorter duration of therapy, and decreased toxicity [18].

The superiority of VDC/IE over VIDE was established in an open-label phase III trial (EE2012) conducted in Europe [18]. In this study, 640 patients with ES were randomly assigned to initial treatment with VDC/IE administered on an interval-compressed schedule (every 14 days with hematopoietic growth factor support) or the VIDE regimen. At a median follow-up of 47 months, VDC/IE improved event-free survival (EFS) and OS compared with VIDE (three-year EFS 67 versus 61 percent; hazard ratio [HR] 0.71, 95% CI 0.55-0.92; three-year OS 82 versus 74 percent; HR 0.62, 95% CI 0.46-0.85). While grade ≥3 toxicity rates were similar between the two treatment arms (90 versus 91 percent), the risk of grade ≥3 febrile neutropenia was lower for VDC/IE compared with VIDE (58 versus 74 percent). The duration of therapy for VDC/IE was also shorter by approximately two months. Based upon this finding, interval-compressed VDC/IE is a standard regimen.

Combining additional chemotherapy agents with VDC/IE does not give any additional clinical benefit in patients with localized disease. In a randomized phase III Children's Oncology Group (COG) trial (AEWS1031) conducted in 629 patients with treatment-naïve localized ES, the addition of vincristine/topotecan/cyclophosphamide to the interval-compressed VDC/IE regimen did not improve EFS or OS [19].

Interval-compressed VDC/IE — In patients <18 years with localized ES, interval-compressed therapy with VDC/IE given every 14 days with hematopoietic growth factor support (table 1) is preferred to administration every 21 days because it improves EFS and OS without increased toxicity or higher risk of second (subsequent) malignant neoplasms.

A randomized phase III trial (AEWS0031) from the COG conducted in North America established interval-compressed therapy with alternating cycles of VDC/IE as the preferred initial chemotherapy regimen for children with localized ES (table 1) [20-22]. In this study, 587 patients with localized ES were randomly assigned to receive 14 alternating cycles of VDC/IE every 21 or 14 days. Most patients enrolled were age <18 years (88 percent). Interval-compressed VDC/IE given every 14 days improved 10-year EFS (70 versus 61 percent) and 10-year OS (76 versus 69 percent) compared with VDC/IE given at 21-day intervals.

In subgroup analyses, compared with chemotherapy every 21 days, interval-compressed VDC/IE every 14 days also improved 10-year EFS in patients <18 years of age (73 versus 64 percent; HR 0.71, 95% CI 0.52-0.99), and in patients with:

●Pelvic primary tumors (67 versus 43 percent)

●Tumor volume ≥200 mL (74 versus 46 percent)

●Viable tumor at the time of surgery (75 versus 47 percent)

The toxicity of both regimens were similar [21]. There were similar rates in the 10-year cumulative incidence of second (subsequent) malignant neoplasms between treatment every 14 days (3.2 percent) and treatment every 21 days (4.2 percent).

Although interval compression is recommended, there is no role for dose escalation. In a randomized phase III trial conducted by the COG, dose escalation of VDC/IE without hematopoietic cell support did not improve outcomes in patients with newly diagnosed localized disease [23]. Furthermore, concerns for an increased risk of secondary malignancies in patients receiving dose-intense therapy have tempered enthusiasm for this approach.

Limited role for high-dose chemotherapy with HCT — The role of consolidation with high-dose chemotherapy with autologous hematopoietic cell transplantation (HCT; rescue) is uncertain and should be reserved for patients enrolled on clinical trials.

Although this approach improved outcomes for children with high-risk localized disease who were treated with the VIDE induction regimen [23], high-dose chemotherapy with autologous HCT has not been used with interval-compressed VDC/IE chemotherapy [24]. Additionally, in a phase III trial (EE2012), interval-compressed VDC/IE improved OS over VIDE among the subgroup of patients with high-risk localized disease (tumor volume ≥200 mL at diagnosis) [18]. It is not known whether high-dose chemotherapy with autologous HCT would lead to further improvements in this population. (See 'Interval-compressed VDC/IE' above.)

The role of consolidative high-dose chemotherapy followed by autologous HCT for localized high-risk disease was evaluated in the European Ewing Tumour Working Initiative of National Groups (Euro-EWING) 99/EWING 2008 trial [23]. High-risk disease was defined as a poor histologic response after receiving six cycles of VIDE (76 percent of the cohort) or tumor volume at diagnosis ≥200 mL if unresected, initially resected, or resected after radiation therapy (RT). In this study, 240 patients with high-risk localized ES were treated with six courses of VIDE plus one course of consolidation vincristine, dactinomycin, and ifosfamide (VAI). Patients were then randomly assigned to either one course of busulfan plus melphalan followed by autologous HCT or seven courses of standard chemotherapy with VAI. At a median follow-up of 7.8 years, high-dose chemotherapy followed by HCT improved both EFS (eight-year EFS 61 versus 47 percent) and OS (eight-year OS 65 versus 56 percent).

Severe acute toxicities were more common in the high-dose chemotherapy group, and three patients died in this group, two of treatment-related toxicity; the third patient did not receive dose-intense chemotherapy or high-dose chemotherapy because of kidney impairment. The risk of secondary malignancies in long-term survivors was not reported.

Patients 18 years or older — For patients ≥18 years with localized ES, we recommend neoadjuvant chemotherapy with alternating cycles of VDC/IE rather than VIDE.

Support for the use of VDC/IE in this population is largely based on extrapolation of data from trials conducted primarily in patients <18 years and described separately (see 'Patients <18 years' above). Compared with VIDE, VDC/IE improves OS, has a shorter duration of therapy, and decreased toxicity. Many of these trials included adults, but adults accounted for a minority of the population.

In patients ≥18 years with localized ES, VDC/IE can be administered as either:

●VDC/IE every 14 days with hematopoietic growth factor support for a total of six cycles (12 weeks of therapy prior to local control), with adjuvant therapy after local control.

●VDC/IE every 21 days with hematopoietic growth factor support for a total of four cycles (12 weeks of therapy prior to local control), with adjuvant therapy after local control.

As discussed above, every-14-day therapy is preferred over every-21-day therapy in younger patients based on the AEWS0031 trial that demonstrated improved EFS and OS in localized ES [20-22]. However, the benefit of every-14-day therapy is less clear in older patients and may be more challenging to administer. Even in the pediatric cohort, the mean cycle duration given was 17 days instead of the goal of 14 days [21]. In addition, while EFS was numerically higher with every-14-day in 31 patients >18 years (10-year EFS of 53 versus 37 percent; HR 0.75, 95% CI 0.38-1.47), it did not reach statistical significance [22]. Similarly, in the EE2012 trial, the effect of interval-compressed VDC/IE was less pronounced in those ≥14 years [18]. While a 14-day cycle may be a laudable goal for treatment of ES in an adult population, the choice of cycle length is made based on patient specific factors including patient age, comorbidities, and the practicalities of the tolerance of treatment.

Local treatment — Following neoadjuvant systemic therapy, patients receive local treatment with surgery, RT, or both. Modern treatment protocols emphasize surgery for optimal local control. However, the choice among these options should be made by a multidisciplinary team and is based on patient characteristics, location of disease, expected toxicity, functional outcomes, potential complications of local treatment, and patient preference.

For most patients, chemotherapy is combined with local therapy that uses one modality (RT or surgery); the combination of all three modalities (RT, chemotherapy, and surgery) is not routinely recommended. However, in a review of local treatment modalities used on the Ewing 2008 study, patients with a poor response to chemotherapy appeared to benefit from the combination of surgery and radiation over surgery alone (EFS HR 0.49, 95% CI 0.27-0.89) [25]. By contrast, the overall group had no difference in EFS, OS, or local relapse amongst the surgery group compared with the surgery and radiation group.

Because no randomized trial has directly compared surgery versus RT, only a relative comparison can be made from retrospective reports and the prospective trials that have mainly tested different multiagent chemotherapy regimens. In many retrospective series, rates of local control and survival are superior after surgery compared with RT alone [2,5,12,26-40]. However, selection bias likely accounts for at least some of these results. Smaller, more favorably situated peripheral tumors are more often resected while larger, axial lesions (which have a higher rate of local failure in most series and a poorer prognosis overall [2,5,26,41]) are more likely to be referred for RT. By contrast, some series suggest local control rates with RT and surgery are similar when controlling for age and primary tumor site [42-44]. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Tumor site and size'.)

Timing of local control — Local treatment (surgery, RT, or both) should be initiated immediately after completion of 12 weeks of induction chemotherapy. An analysis of data derived from the National Cancer Database noted that patients initiating local therapy by week 15 versus week ≥16 had improved five-year OS (78.7 versus 70.4 percent) and 10-year OS (70.3 versus 57.1 percent) [45]. On subgroup analysis, the difference in OS according to time to local therapy was particularly large in patients receiving RT alone (five-year OS 80.0 versus 58.9 percent). (See 'Multimodality therapy for localized disease' above.)

Resectable disease

Surgery preferred — Surgery is the preferred modality for tumors that can be resected with adequate margins, and acceptable morbidity and functional outcome. The determination about resectability of a lesion is often made by a multidisciplinary team. Tumor location, size, and impact of adjacent structures all influence the decision about whether a tumor should be resected. Tumors arising in dispensable bones (eg, fibula, rib, small lesions of the hands or feet) are generally able to be resected with minimal morbidity, whereas tumors of the long bones of the leg, distal humerus, or ulna are typically resected using intercalary techniques (allografts, autografts, or metallic prostheses) or joint replacement. In addition, ES is a chemotherapy-sensitive disease and lesions that may initially appear unresectable may be amenable for resection after neoadjuvant chemotherapy. (See "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management".)

Modern surgical techniques offer the opportunity for limb reconstruction with endoprostheses that can preserve function and minimize disability. Patient selection is evolving as surgical morbidity has decreased with advances in reconstruction and prosthesis options (eg, 3D printed, custom prostheses) [46-48]. For larger, bulky tumors that are resectable or borderline resectable, sometimes preoperative RT is offered to help facilitate the resection. The surgical principles that apply to resection of the primary tumor and reconstruction are similar to those in patients with osteosarcoma and are discussed elsewhere. (See "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management".)

Support for surgery as the preferred modality is based on the following:

●It avoids the risk of secondary radiation-associated sarcomas.

●It allows for analysis of tumor necrosis in the excised tumor to refine prognosis estimates.

●It avoids the risk of impacting bone growth and resulting deformity in the skeletally immature child.

The COG reported improved rates of local control in patients treated with surgery. This study analyzed local control outcomes in patients treated with surgery, radiation, or both on three consecutive protocols using standard-dose, five-drug chemotherapy every 21 days [38,49]. Local failure rates were lower among patients treated with surgery alone or surgery plus RT when compared with those treated with RT alone (3.9 and 6.6 versus 15.3 percent, respectively). On multivariate analysis, compared with surgery, use of RT was associated with a higher risk for local failure (HR 2.41, 95% CI 1.24-4.68), although there were no significant differences detected in EFS (HR 1.42, 95% CI 0.94-2.14), OS (HR 1.37, 95% CI 0.83-2.26), or distant failure (HR 1.13, 95% CI 0.70-1.84). As to be expected, patients who underwent surgery were younger and had more appendicular primary tumors. The local failure incidence was higher for pelvic tumors (13.2 percent) than for other sites (9.1 percent for extraskeletal tumors, 5.4 percent for extremity tumors, 5.3 percent for axial non-spine tumors, and 3.6 percent for spine tumors).

Similarly, a secondary analysis of the Ewing 2008 trial found that the adjusted hazards for any event (a composite endpoint of EFS, OS, and local recurrence) was increased for those treated with RT alone compared with surgery (HR 1.53, 95% CI 1.02-2.31) [25]. The analysis included 863 patients who received surgery alone, surgery and radiation, or radiation alone while receiving protocol-specified chemotherapy. When compared with other locations, pelvic and spine primaries were more frequently treated with RT as the primary local treatment modality (40 and 31 percent). This likely reflected the inability to resect tumors in these locations.

Although a benefit for intraoperative RT (IORT) has been suggested in retrospective series involving a small number of patients [50,51], this approach is not commonly used. Peripheral nerves are dose-limiting structures for IORT, so the decision to use IORT must take into account the risk of severe neuropathy and soft tissue necrosis.

Bulky tumors in difficult sites — For bulky tumors in difficult sites such as the pelvis, decisions about resectability and using RT are made by a multidisciplinary team.

For lesions that are felt to have a high likelihood of positive margins, definitive RT is often pursued. However, combined surgery and RT might allow for a more limited surgical procedure, better functional outcome, and enhanced local control as compared with single-modality therapy [26,52-54]. Preoperative RT is considered when it is likely that margins will be positive. This may shrink the tumor and allow for resection. However, it can also adversely impact the soft tissues making dissection more challenging and cause a greater likelihood of wound complications. Postoperative radiation is also an option, but can be challenging to identify where to direct RT if there is a large field and uncertainty exactly where the positive margin was located.

Unresectable disease — RT is preferred over surgery when the tumor is unresectable or if a resection will result in significant morbidity. The following are examples of locations that are typically treated with RT instead of surgery:

●Primary tumors of the spine [55].

●Tumors of the skull or facial bones because of the difficulty in achieving negative margins without substantial functional deficit [56].

●Tumors in locations such as the periacetabular region of the pelvis or the sacroiliac joint because of a loss of critical function from surgical procedures that involve resecting the acetabulum or resecting sacral nerve roots. However, the treatment of pelvic tumors is controversial. Lesions of the iliac wing, ischium, or pubis can usually be resected after a good response to chemotherapy without substantial functional morbidity [28,52,57-59]. (See "Bone sarcomas: Preoperative evaluation, histologic classification, and principles of surgical management".)

Support for the use of RT comes from case series or secondary analyses of trials that were prospectively evaluating the use of chemotherapy. This approach has limitations as many patients who were referred for RT had lesions that were difficult to resect. The results are summarized as below:

●No difference in local failure or EFS rates were shown in a subgroup of 75 patients with nonmetastatic pelvic ES treated with surgery, RT, or both [60]. The Intergroup study 0091, a randomized comparison of vincristine, doxorubicin, cyclophosphamide, and dactinomycin with or without alternating IE (ifosfamide plus etoposide), allowed treating clinicians to choose the local control modality. Five-year EFS was 42 percent in patients who received surgery, 52 percent for those receiving RT, and 47 percent in patients who received both treatment modalities.

●However, the outcomes for surgery versus RT may depend on site of disease. In a COG analysis of patients on trials comparing chemotherapy protocols who were treated with surgery or RT, local failure rates with surgery were numerically lower but did not reach statistical significance for the subgroup of patients with axial nonspine (2.5 versus 10.6 percent), spine (0 versus 5.6 percent), and extraskeletal tumors (7.7 versus 10.9 percent) [49]. However, this review, which included 956 patients treated with ifosfamide- and etoposide-based chemotherapy on three consecutive protocols, did show lower rates of local failure in patients with extremity and pelvic tumors treated with surgery compared with those treated with RT. Local failure rates did not differ based on location for surgically treated patients, but did for patients treated with RT.

●Adults with ES who are treated with modern chemotherapy protocols and RT techniques have outcomes that are similar to those of younger patients, with anticipated local control rates of approximately 80 percent after definitive RT [61-63]. With a median follow-up of 41 months, disease-free survival and OS rates for patients with nonmetastatic disease at presentation were 70 and 86 percent, respectively. [61] There were nine local recurrences, five of which developed in patients who presented with overt metastases. The three-year actuarial local control rates for the entire group and those undergoing definitive RT alone were 77 and 78 percent, respectively. The presence of overt metastases emerged as the only significant adverse prognostic factor for local control, a finding that has been confirmed by others [26,64]. Long-term treatment-related complications were minimal, and there were no secondary malignancies.

●In patients with pelvic primaries treated on the Euro-EWING99 trial, an analysis of local control methods found no association with OS at five years in patients treated with definitive radiation versus surgery and radiation (73 versus 78 percent) [65].

Suboptimal RT delivery techniques with inadequate tumor coverage may have contributed to poorer outcomes in early series. As an example, in the Cooperative Ewing Sarcoma Study (CESS)-81 trial, relapse-free survival (RFS) for patients undergoing irradiation of the primary site in the initial phase of the trial was only 55 percent, significantly lower than for surgically treated patients [30]. On review, the high rate of local failure after RT alone (50 percent) was attributed to geographic "miss." After a quality assurance program was initiated, local control rates improved substantially and RFS increased to 80 percent. In a subsequent study, CESS-86, which included central quality assurance, the five-year RFS rates following chemotherapy plus either RT or surgery were 67 and 65 percent, respectively [66].

Treatment volume — Modern treatment protocols for ES include tailored RT fields that target the site of the primary lesion with a margin rather than the whole bone [67]. While whole bone RT had been used in past, a randomized trial of whole bone versus tailored-field RT after 12 weeks of induction chemotherapy was stopped early for benefit after demonstrating that a tailored field was as effective as whole bone RT [6]. At three years, the rates of local control and EFS were 76 and 54 percent, respectively. Additional studies have shown the benefit of reducing the irradiated field and targeting higher doses to the site of the initial primary tumor [30,64].Tailored RT is administered to an initial clinical target volume that includes the original bone and soft tissue tumor extent with a 1 cm margin. For patients with chest wall tumors and malignant effusions, the field includes the entire ipsilateral hemithorax to a dose of 15 Gy prior to cone-down to the chest wall tumor. For patients where the soft tissue tumor displaces but is not thought to violate an anatomic compartment (ie, pelvic ES displacing bowel or bladder), the initial target volume includes the post-chemotherapy soft tissue extent of disease abutting the anatomic compartment, rather than the original prechemotherapy extent of disease.

The initial clinical target volume is treated to 45 Gy in 25 fractions followed by a field reduction to encompass the postchemotherapy gross soft tissue tumor as well as all of the originally involved bone with a margin of 1 cm; this volume is treated to another 10.8 Gy in six fractions in patients undergoing definitive RT and to 5.4 Gy in three fractions for patients undergoing adjuvant RT for microscopic residual disease after surgery (with boost up to 10.8 Gy for any sites of gross disease after surgery). Additional margin expansions on these volumes are used for the planning target volume expansion to account for any variability in daily setup; this margin should be individualized by the treating clinician based on the reproducibility of immobilization and the kind of image guidance used but is generally a minimum of 5 mm and usually in the range of 0.5 to 1 cm.

Attention to potential RT effects on normal tissue is critical in radiation planning to minimize late effects, particularly in children:

●Treatment of uninvolved epiphyseal growth plates is avoided to minimize treatment-induced limb shortening [68].

●Circumferential irradiation of a limb is avoided to reduce the risk of limb edema and fibrosis.

●Gonadal avoidance or additional shielding (for the testes) is important to retain fertility.

●Excellent results can be achieved from RT of lesions of the hand or foot [67]. However, walking places repetitive stress on the soft tissues of the sole of the feet, and dose sparing for the sole of the foot can improve the functional outcome of treatment.

●Irradiation of the Achilles tendon is usually avoided. Lesions of the calcaneus can be treated; although some atrophy of the heel pad has been reported, the functional impairment is usually minor [69].

●Nail beds should be excluded from the radiated field when possible.

●For pelvic tumors, distention of the bladder prior to each day's treatment can reduce the amount of small bowel in the radiated field, but care must be taken to avoid radiation to significant portions of the bladder in patients receiving cyclophosphamide or ifosfamide to reduce the risk of hemorrhagic cystitis.

Radiation modality — RT for ES can be delivered using three-dimensional conformal radiation therapy (3D-CRT), intensity modulated radiation therapy (IMRT), or proton therapy. The choice of which modality to use is determined by location of the lesion, potential long-term impact of treatment, and availability of different modalities. It is anticipated that the improvements in RT techniques will reduce the risk of long-term complications while maximizing local control rates.

●IMRT versus 3D-CRT – IMRT is typically preferred over 3D-CRT. IMRT utilizes variable, computer-controlled intensities within each RT beam, in contrast to the uniform doses within each 3D-CRT beam. Accordingly, compared with 3D-CRT, IMRT typically achieves better dose conformity to the planned target and better sparing of normal tissue, thereby providing high rates of local control and a lower incidence of some late complications. [53]. IMRT is especially useful when target volumes have complex shapes or concave regions. (See "Radiation therapy techniques in cancer treatment", section on 'Intensity-modulated radiation therapy'.)

The tradeoff of improved dose conformity with IMRT is the delivery of low-moderate doses of radiation to a larger volume of surrounding tissue than with 3D-CRT techniques. In addition, patients receiving IMRT may also receive a larger total-body dose because of leakage of radiation [70]. Because of this, there is a theoretical concern for a higher potential risk of second malignancy with IMRT, but clinical studies have not shown this to be true. One observational study of adult and pediatric patients treated for nine tumor types reported that second malignancy rates for IMRT were comparable with those with 3D conformal photons but were reduced with protons (relative risk 0.31, 95% CI 0.26-0.36) [71]. These questions will be further clarified by additional follow-up of patients treated with IMRT.

●Proton beam therapy — Compared with photon beam irradiation, proton beam therapy permits the delivery of high doses of RT to the target volume while reducing the radiation dose received by normal tissues distal to the target. Depending on the tumor target configuration and the portal selection, this can result in up to 60 percent reduction in the dose received by uninvolved normal tissue compared with other radiation methods [71]. The ability of proton beam irradiation to minimize exit dose, renders it useful for tumors located close to sensitive tissues. For example, it may be particularly beneficial for ES arising in the bones of the paranasal sinus (image 1 and image 2), the pelvis, and the spine and for those requiring higher doses. (See "Radiation therapy techniques in cancer treatment", section on 'Particle therapy'.)

Proton beam therapy has demonstrated better sparing of the intestine, rectum, bladder, pelvic bone marrow, and femoral head as compared with conventional photon irradiation for pelvic tumors [72]. In a retrospective review of 30 children treated with proton therapy for ES with a median follow-up of 38 months, three-year actuarial rates of EFS, local control, and OS were 60, 86, and 89 percent, respectively [73]. Proton beam therapy was well tolerated with only mild to moderate skin reactions. There were four cases of late hematologic malignancies, but these are known risks of chemotherapy used for ES.

Since proton beam therapy has a lower volume of normal tissue in the irradiated field, it may have a lower risk of secondary in field malignancies [74,75]. A review from the National Cancer Database of pediatric and adult patients treated for nine tumor types suggested a reduction in radiation-associated second cancers with protons compared with photons (relative risk of 0.31, 95% CI 0.26-0.36) [71]. While there is still concern about late radiation effects including neutron scatter radiation, modern treatment protocols have decreased this risk [70,76,77]. Further study is needed to confirm these results [78-80].

Dose — For patients treated with chemotherapy and RT, cooperative studies in the United States have employed 45 Gy in 25 fractions to the initial clinical target volume as defined above, followed by a 10.8 Gy boost in six fractions to the site of original bony disease and any residual soft tissue disease that remains following chemotherapy. Because of the proximity of the spinal cord, the dose to vertebral body primaries has been often limited to 45 Gy, although techniques such as proton beam therapy, IMRT, and stereotactic body RT can permit the safe delivery of higher doses [81].

RT dose is an important factor in local control, particularly for large tumors [64,82]. While 45 Gy is typically used, lower doses (eg, 30 to 36 Gy) have also been studied [82,83]. In one early report of patients undergoing definitive RT, local control rates were 90 percent using RT doses of 35 Gy for smaller lesions (≤8 cm) with a favorable response to chemotherapy, but they were inferior (52 percent) for larger lesions [82].

Dose-escalation RT to limited fields may be beneficial for high-risk tumors and may reduce local recurrence rates, although this approach is considered investigational. A single-institutional phase II trial employed focal limited-margin RT using conformal or intensity-modulated techniques [84]. Unresected tumors <8 cm at diagnosis received a standard dose of 55.8 Gy, and tumors >8 cm received an escalated dose to 64.8 Gy. Patients with microscopic residual disease after resection received adjuvant RT to 50.4 Gy. Adjuvant brachytherapy was permitted in selected patients. Forty-five patients were enrolled, 26 with localized and 19 with metastatic disease. Seventeen patients received adjuvant RT, 16 received standard-dose, and 12 received escalated-dose RT. Failures included 1 local, 10 distant, and 1 combined local/distant. The estimated 10-year cumulative incidence of local failure was 4.4 ± 3.1 percent, with no statistical difference seen between RT treatment groups; there were no local failures in the escalated-dose RT treatment group.

By contrast, a randomized phase III trial of 95 patients with localized ES demonstrated improved five-year local control in the dose-escalation arm compared with standard dose RT (76 versus 49 percent) [85]. However, the OS did not achieve statistical significance at five years despite a numerically higher value (59 versus 45 percent). There was a higher incidence of skin toxicity in the dose-escalation arm compared with standard RT (10 versus 2 percent). Similarly, a retrospective study of 32 patients with newly diagnosed Ewing sarcoma ≥8 cm treated with dose escalation (59 to 69 Gy) was associated with high rates of local control (five-year local failure rate of 6.6 percent) [86].

This approach, if validated, has particular appeal because of the reduction in radiation target volume, which should reduce both acute and late toxicity, as well as the selective use of dose escalation for high-risk lesions >8 cm, reported in multiple series to be at higher risk for local failure with conventional RT doses. Another report also suggests that patients treated with definitive radiation doses ≥56 Gy had a lower incidence of local recurrence (17 versus 28 percent) [87].

Radiation schedule — Conventional RT schedules consist of once daily RT doses of 1.8 Gy per fraction and this approach is the standard of care for ES. Accelerated fractionation RT does not seem to improve rates of local control or survival for ES [29,30,53,66,88].

A case series of 75 patients with ES suggested limited field sizes with hyperfractionated RT (1.2 Gy twice daily with a six hour interfraction interval) could minimize long-term complications and provide superior functional outcomes [29,89]. However, treatment with a twice daily approach is not standard and should only be considered as part of a clinical trial.

Adjuvant therapy — Since most treatment failures are attributable to systemic metastatic disease, local therapy considerations should be planned to avoid complications that might compromise the administration of systemic therapy. After patients have healed from surgery, we restart systemic therapy as soon as possible. For patients treated with radiation rather than surgery, chemotherapy is given concurrently with radiotherapy, with the exception of doxorubicin which is held during radiotherapy.

Adjuvant chemotherapy — For patients receiving VDC/IE, we typically administer a total of 14 to 17 cycles which includes those cycles received as neoadjuvant therapy. A total of either 14 or 17 cycles is acceptable as both regimens have been used in large trials and both have shown favorable outcomes [19,21]. However, they have not been directly compared. A decision about the total number of cycles must take into account patient specific factors such as tolerability of treatment and ability to complete the total number of cycles (algorithm 1).

The routine administration of intensive multiagent chemotherapy, which can eradicate metastatic deposits, has had a dramatic impact on outcomes. For ES, from 1975 to 2010, the five-year survival rate increased from 59 to 78 percent for children younger than 15 years and from 20 to 60 percent for adolescents aged 15 to 19 years [90,91].

When to use adjuvant radiation therapy — For patients with residual disease after surgery defined as either positive margins or gross residual disease, we suggest adjuvant RT to improve local control. Adjuvant RT reduces the risk of local failure, although effective chemotherapy can also reduce this risk [3,9]. By contrast, adjuvant RT is not usually offered to patients without residual disease (ie, resected with negative margins) to avoid exposing them to the risk of a radiation-induced malignancy.

Indications for adjuvant RT include:

●Residual microscopic or gross disease after surgery, or inadequate surgical margins (ie, a marginal or intralesional resection (table 2)).

●Adjuvant hemithorax irradiation is indicated in patients with high-risk chest wall primary tumors (eg, with cytology-positive initial pleural effusion, pleural infiltration, and intraoperative contamination of the pleural space).

Positive or inadequate margins or gross residual disease — For patients with positive margins or gross residual disease following surgery, we suggest adjuvant RT. Patients who are left with macroscopic amounts of viable tumor in the resected specimen following neoadjuvant chemotherapy have worse survival than those with minimal or no residual tumor [2,27,42,92-96]. In addition, inadequate resection margins (ie, a marginal or intralesional resection (table 2)) are associated with a worse outcome as compared with radical or wide resection [34,97].

For patients undergoing adjuvant postoperative RT, the dose depends on the amount of residual tumor with doses of 50.4 Gy administered for microscopic disease (positive margins), and 55.8 Gy for gross residual disease. Although some studies suggest that doses as low as 30 Gy may be sufficient in the adjuvant setting [83], further studies are needed before lower doses of RT can be accepted as standard practice.

The Euro-EWING group reported a large observational study of 599 patients who underwent surgery after chemotherapy for standard-risk localized disease [98]. Postoperative RT was recommended for patients with inadequate surgical margins, and it was also common practice for patients with a vertebral tumor or with a pleural effusion contiguous to a rib or thoracic wall primary tumor. After controlling for possible confounders, there was a reduction in local recurrence in patients treated with surgery plus postoperative RT compared with surgery alone (HR 0.43, 95% CI 0.21-0.88). They also observed a nonsignificant trend for benefit associated with postoperative RT for disease-free survival, EFS, and OS. Among the 132 patients for whom RT would have been recommended according to the policies of the Euro-EWING group, 50 did not receive it. Among these patients, local recurrence rates were 30 percent at eight years compared with 11 percent when RT was performed.

On the other hand, a lack of benefit for adjuvant RT was suggested in a single-institutional retrospective series of 512 patients, in which the addition of moderate-dose adjuvant RT (45 Gy) to patients with inadequate margins did not seem to improve local control or overall disease-free survival [99]. The RT dose used in this series is on the low end of the range currently used for microscopic residual disease, and lower than that used for macroscopic residual disease and may have contributed to the lack of difference seen.

Prophylactic bilateral whole-lung irradiation — The role of adjuvant bilateral whole-lung irradiation is unsettled and we do not recommend it as prophylactic treatment in the setting of localized disease. The IESS-I trial showed that bilateral whole-lung irradiation was an effective adjuvant treatment for patients with localized ES, but prolonged follow-up favored four-drug multiagent chemotherapy [4]. Prophylactic whole-lung irradiation has not been studied in any subsequent trial of localized disease, and no studies have used it in addition to modern combination chemotherapy regimens [99], except in patients with lung metastases at diagnosis.

TREATMENT FOR METASTATIC DISEASE — 

Patients with overt metastatic disease at presentation have a worse outcome than those with localized disease. However, aggressive multimodality therapy can relieve pain and delay progression.

In a review of 13 different series in which patients with metastatic ES were predominantly treated with chemotherapy, five-year event-free survival (EFS) and overall survival (OS) rates averaged 25 (range 9 to 55) and 33 (range 14 to 61) percent, respectively [1,5,100-110]. The small numbers of patients in each series and the heterogeneity in location and extent of metastatic disease probably account for these wide variations in outcome. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Disease extent'.)

A subset of patients with advanced disease may be cured by multimodality therapy, although the long-term survival rates are clearly lower than for patients with localized disease [111]. Because it can be difficult to predict which patients with metastatic disease will be long-term relapse-free survivors [101], treatment for select patients with potentially favorable metastatic disease should be administered with curative intent. Clinicians experienced in the treatment of ES must direct multimodality therapy and coordination between specialists [5,30,82]. The site(s) of metastatic disease is an important variable. For children with isolated lung and pleural metastases, five-year EFS rates up to 50 percent are reported with multimodality therapy; for metastases involving bone or bone marrow, EFS rates fall to 10 to 20 percent; and, for combined sites, to less than 15 percent [108].

Initial selection of chemotherapy

VDC/IE for metastatic disease — Patients with disseminated disease at diagnosis often respond well to the same type of systemic chemotherapy as is used for localized disease. Randomized trials are limited in this population as it accounts for only 25 to 30 percent of patients with ES. Whenever possible, patients with newly diagnosed metastatic ES should be offered enrollment in clinical trials evaluating novel approaches. In the absence of a clinical trial, regimens similar to those used for the treatment of patients with newly diagnosed localized disease are commonly used. (See 'Neoadjuvant chemotherapy' above.)

For most patients with metastatic ES, we suggest alternating cycles of vincristine/doxorubicin/cyclophosphamide and ifosfamide/etoposide (VDC/IE) based on extrapolation of data from trials in localized disease. In contrast to the experience for patients with localized disease, studies in patients with metastatic disease have not demonstrated a specific benefit for the addition of I/E to the VDC backbone at the time of diagnosis [103,104,112-115]. As such, some centers utilize VDC for patients with unequivocal evidence of metastatic disease. Given robust improvement in survival with IE in the localized setting, patients with equivocal findings should be treated with VDC/IE.

Several trials have included interval-compressed VDC/IE administered every 14 days in the context of metastatic ES [18,116]. As an example, the COG AEWS1221 trial evaluated the addition of ganitumab prescribed interval-compressed VDC/IE for all patients, with outcomes similar to the historic experience [116]. Similarly, the EE2012 trial described above included patients with localized or metastatic disease and showed a benefit for interval-compressed VDC/IE compared with vincristine, ifosfamide, doxorubicin, and etoposide (VIDE) chemotherapy [18]. The interval-compressed VDC/IE regimen is frequently used in this context as it does not appear to be more toxic and patients complete therapy more quickly. However, given the lack of randomized data comparing every-14-day versus every-21-day chemotherapy in this population, an acceptable alternative is for administration of chemotherapy every 21 days. (See 'Interval-compressed VDC/IE' above.)

By contrast, data from a nonrandomized trial of dose intensification of VDC/IE with augmented alkylator doses did not suggest a benefit in patients with newly diagnosed metastatic ES [117].

Limited role for high-dose chemotherapy in metastatic disease — High-dose chemotherapy with hematopoietic cell transplantation (HCT; rescue) does not have a significant role in the treatment of patients who present with metastatic disease, and we do not offer this approach. In randomized studies, the addition of high-dose chemotherapy with HCT did not significantly improve EFS or OS and was associated with significant toxicity [118-120].

High-dose chemotherapy with HCT was evaluated in a randomized clinical trial (Euro-EWING 99 and EWING 2008) [118]. In this trial, 287 patients with isolated pulmonary (lung or pleural) metastatic disease received six cycles of VIDE and one cycle of vincristine, dactinomycin, and ifosfamide (VAI). Subsequently, patients were randomly assigned to receive either one course of busulfan plus melphalan high-dose chemotherapy followed by autologous HCT, or seven cycles of conventional chemotherapy with VAI followed by whole-lung irradiation. At a median follow-up of approximately eight years, there was no statistically significant difference in EFS between the two groups (eight-year EFS 53 versus 43 percent; hazard ratio [HR] 0.79, 95% CI 0.56-1.1), and OS results were comparable (eight-year OS 55 versus 54 percent; HR 1, 95% CI 0.7-1.44). Additionally, rates of infection as well as gastrointestinal and liver toxicities were higher in the high-dose chemotherapy group and included four deaths versus no deaths in the conventional chemotherapy group.

Similar results were demonstrated in a separate open-label phase III trial (Ewing 2008R3) in which 109 patients with metastatic ES (excluding those with pulmonary metastases only) were treated with six cycles of VIDE and consolidation therapy with eight cycles of vincristine, actinomycin D, and cyclophosphamide [120]. Patients were then randomly assigned to either treosulfan plus melphalan high-dose chemotherapy followed by HCT or no further treatment. At median follow-up of 3.3 years, the addition of high-dose chemotherapy with HCT did not improve three-year EFS (21 versus 19 percent; HR 0.85, 95% CI 0.55-1.32). However, in a posthoc analysis of the subset of 41 patients less than 14 years of age, a potential EFS benefit for this approach was demonstrated (three-year EFS 39 versus 9 percent; HR 0.4, 95% CI 0.19-0.87). Given the limited sample size and lack of preplanned analysis for this subgroup, these results should be interpreted with caution.

Studies evaluating HCT in patients with localized disease are presented separately. (See 'Limited role for high-dose chemotherapy with HCT' above.)

Local control in patients with metastatic disease — Outcomes are best when chemotherapy is combined with optimal local therapy, including radiation and sometimes resection of sites of gross metastatic disease [109,121-123].

Local control of the primary tumor — Most patients benefit from surgery and/or radiation therapy for local control of the primary tumor. Management of the primary site in the setting of metastatic disease requires multidisciplinary input. Local treatment of the primary lesion is important to prevent complications such as fracture, pain, or wound development. In addition, combining local treatment with chemotherapy can increase overall response rates, which will hopefully translate into improved long-term outcomes. However, long-term relapse-free survival is still only achieved in the minority of patients [101].

The choice between surgery versus radiation therapy (RT) for local control is patient specific. As an example, we typically offer surgical resection of the primary to patients who have experienced a significant volume reduction in the primary site after chemotherapy and only have small-volume metastatic disease that is amenable to either resection or RT. By contrast, the presence of diffuse metastatic disease can make it difficult to justify a large resection, which would necessitate a lengthy period off systemic chemotherapy [52,101]. As such, RT is typically used to treat the primary tumor in this setting and usually provides adequate local control with acceptable morbidity [5,30,66,82]. However, if substantial amounts of bone marrow will need to be included in the radiation treatment volume and/or the patient's performance status is poor such that the potential risks of RT outweigh the potential benefits, RT to the primary site may be deferred.

Isolated pulmonary metastases — For patients with pulmonary metastases, we offer a multimodality approach that includes chemotherapy and supplemental low-dose whole-lung irradiation. Whole-lung radiation is typically administered after completion of planned systemic therapy. Surgical resection is reserved for lung metastases that do not resolve with chemotherapy [124]. Patients with lung metastases have a better prognosis than patients with metastatic disease at other sites (ie, bone or bone marrow). Although complete resection may be possible, it is not sufficient therapy on its own; chemotherapy is a necessary component of therapy, and five-year survival between 20 and 40 percent can be achieved with multimodality therapy [108,125,126].

Retrospective reports from large cooperative groups and single institution series suggest that low-dose bilateral whole-lung irradiation benefits patients with the ES family of tumors presenting with pulmonary metastases, even if all lesions are resected or completely respond to chemotherapy [3,30,104,108,109,122,126-129]. In the Cooperative Ewing Sarcoma Study (CESS) and European Intergroup Cooperative Ewing Sarcoma Study (EICESS) trials, the addition of lung RT reduced the rate of pulmonary relapse and improved EFS (from 19 to 40 percent in EICESS) [104,126]. However, most of the available data supporting benefit from RT in this setting are from nonrandomized series, and patient selection factors confound interpretation of the data. There are no randomized trials that confirm the benefit of this approach.

Despite the lack of controlled trials, low-dose bilateral lung irradiation (15 to 18 Gy, in daily 1.5 Gy fractions) with a focal boost dose to a total of 40 to 50 Gy to large deposits is typically administered for patients with pulmonary metastases who have had a good response to chemotherapy. Recommended doses are 15 Gy in 10 fractions for patients <14 years old and 18 Gy in 12 fractions for those >14 years old [18].In the rare event of a child <7 years, we would give 12 Gy in 8 fractions. Long-term toxicity appears to be acceptable. This was shown in a retrospective analysis of data from the prospectively performed EICESS, in which pulmonary function tests were available for 37 patients who were treated with whole-lung irradiation (with or without a boost) [130]. At a median follow-up of 25 months, an assessment of pulmonary complications in patients treated with whole-lung irradiation without further boost were classified as none (43 percent), mild (29 percent), moderate (21 percent), and severe (7 percent). Slightly higher complication rates were reported in patients who had an additional radiation boost or surgery to the thorax in addition to RT. The addition of consolidation whole-lung RT appears to be well tolerated in adult patients with lung metastases, with minimal reported significant acute or late toxicity and a three-year rate of freedom from pulmonary relapse of 45 percent [129]. Boost using stereotactic body RT is also an attractive option. It is an area of active investigation in patients with ES and oligometastatic sites of disease [131].

Bone and soft tissue metastases — For young, fit patients with solitary or circumscribed bone or soft tissue metastases, we suggest aggressive multimodality treatment. Approximately 10 percent of such patients will achieve long-term survival with this approach. RT can be delivered to metastatic lesions (doses equivalent to 40 to 50 Gy) in addition to irradiation of the primary tumor. One small series of 13 patients with metastatic ES and rhabdomyosarcoma reported only one local failure at irradiated sites with minimal associated toxicity [131]. Stereotactic body RT is an area of active investigation in patients with oligometastatic sites of disease [132].

RECURRENT DISEASE — 

Most relapses occur within two years of initial diagnosis, but late relapse is not uncommon [133-135]. In a report from the Childhood Cancer Survivor Study, the 20-year cumulative incidence of a late recurrence among five-year survivors of ES was 13 percent [135]. For this reason, patients should be followed indefinitely for the potential of late relapse. (See 'Posttreatment surveillance' below.)

In general, survival after an early relapse is poor, with few survivors among those who relapse within two years of therapy [136,137]. By contrast, up to 15 to 20 percent of those who relapse later may survive long-term. Other prognostic factors associated with an increased risk of death in patients with recurrent ES include recurrence at combined local and distant sites, and an elevated lactate dehydrogenase at initial diagnosis. Some patients with a relapse may be candidates for local therapy in addition to systemic therapy. (See 'Local control in patients with metastatic disease' above.)

Initial recurrence — For patients with an initial recurrence of ES, we suggest either high-dose ifosfamide or the combination of irinotecan plus temozolomide (IT) with or without vincristine, rather than topotecan plus cyclophosphamide (TC) [138-140]. For patients who decline or are ineligible for either of these regimens due to potential toxicities, TC is an appropriate alternative. Since treatment intent is generally palliative and each chemotherapy regimen carries a different toxicity profile, clinicians should discuss the risks and benefits of each treatment with their patients. Some patients may be candidates for local therapy. (See 'Local control in patients with metastatic disease' above.)

Although the prognosis for patients with recurrent disease is poor, some patients can be successfully salvaged, particularly patients with late relapses [141,142]. Sites of recurrence, prior treatment, and relapse-free interval affect remaining treatment choices. Most patients with recurrent disease will receive systemic therapy prior to attempts at additional local control measures. Patients relapsing after a lengthy disease-free interval off chemotherapy may respond again to the same agents used as part of initial therapy.

An international phase III trial (rEECur) of 439 patients with recurrent or primary refractory ES randomly assigned treatment to one of four different chemotherapy regimens: high-dose ifosfamide, TC, IT, and gemcitabine plus docetaxel (GD) [143]. During the first two interim analyses, IT and GD were dropped from further study as these regimens were predicted to have a lower probability of superiority compared with the remaining arms.

The phase III portion of rEECur compared high-dose ifosfamide versus TC in 146 patients. At a median follow-up of 40 months, presented in abstract form, high-dose ifosfamide demonstrated [143]:

●Numerically higher event-free survival (EFS; median 5.7 versus 3.5 months; six-month EFS 47 versus 37 percent; hazard ratio [HR] 0.73, 95% CI 0.51-1.05) and overall survival (OS; median 15.4 versus 10.5 months; one-year OS 55 versus 45 percent; HR 0.73, 95% CI 0.50-1.08); this difference, while potentially clinically significant, did not reach statistically significance. Objective response rates were also higher for high-dose ifosfamide compared with TC (30 versus 21 percent). In a subgroup analysis, high-dose ifosfamide conferred a higher EFS benefit versus TC for children (age <14 years) compared with adolescents and adults (age ≥14 years).

●Higher rates of grade ≥3 neurotoxicity (7 versus 0 percent), nephrotoxicity (8 versus 0 percent), and infection (14 versus 8 percent), leading to higher rates of treatment discontinuation. Other grade ≥3 toxicity rates were similar between the two treatment arms including febrile neutropenia (25 versus 26 percent), vomiting (1 percent each), nausea (3 versus 0 percent), and diarrhea (1 percent each).

Due to the trial design of rEECur, it is difficult to compare the efficacy of high-dose ifosfamide versus IT. These two regimens were not directly compared in the randomized phase III portion of the study, and the dropping of the IT treatment arm implies that its efficacy was inferior to high-dose ifosfamide. However, in preliminary results of an interim analysis of rEECur, at median follow-up of nine months, among the 127 patients treated with IT arm, median EFS and OS were 4.7 and 13.9 months, respectively, which were only slightly lower than that seen with high-dose ifosfamide (5.7 and 15.7 months, respectively) but higher than that seen with TC (3.5 and 10.5 months, respectively) [144]. Grade ≥3 toxicity for IT were diarrhea (17 percent), nausea, vomiting (6 percent each), fatigue, and febrile neutropenia (3 percent each). Since data on patient-reported outcomes are limited, toxicity profiles are different for each regimens, and treatment is generally palliative in this situation, both IT and TC remain reasonable options in patients with recurrent ES. [125,145].

For patients treated with IT, a randomized single-center phase II trial reported a higher response rate using a 10-day irinotecan schedule rather than the 5-day schedule studied in rEECur [138]. This regimen also included vincristine and the extent to which vincristine contributed to these outcomes remains unclear. These findings have not been validated in larger studies and choice of the 5-day versus 10-day schedule remains individualized based upon goals of care and feasibility of frequent infusion visits.

Multiply relapsed disease — Patients with multiply relapsed ES are treated with palliative intent. Options include:

●Multitargeted tyrosine kinase inhibitors – The multitargeted tyrosine kinase inhibitors cabozantinib and regorafenib are alternative options for subsequent-line therapy in patients with relapsed/refractory or metastatic disease. Cabozantinib has demonstrated activity in patients with advanced ES, with an objective response rate of 26 percent and median progression-free survival of five months in one phase II trial (CABONE) [146]. Likewise, a non-randomized phase II trial of regorafenib reported a 10 percent response rate in patients with recurrent ES [147]. In addition, a randomized placebo-controlled phase II trial of regorafenib reported nominally longer PFS in patients randomized to regorafenib compared with placebo [148].

●Investigational agents – Patients with recurrent or advanced ES should be encouraged to participate in clinical trials, where available. A full listing of clinical trials in this patient population is available at clinicaltrials.gov. Future therapies will likely emerge as the fundamental biology of the fusion oncoproteins driving this disease is better understood [111,149]. (See "Epidemiology, pathology, and molecular genetics of Ewing sarcoma" and "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Prognostic factors'.)

In patients with recurrent or refractory ES, initial studies suggested activity for antibodies that target insulin-life growth factor-1 receptor, such as ganitumab [150-154]. However, in a randomized phase III trial of patients with treatment-naïve metastatic ES, the addition of ganitumab to interval-compressed chemotherapy with vincristine/doxorubicin/cyclophosphamide and ifosfamide/etoposide did not improve EFS and increased toxicity [116].

Local therapies — Management of a local recurrence usually includes surgery (and possibly an amputation if the local recurrence involves an irradiated extremity), radiation therapy (RT), or both. RT to bone lesions usually provides pain relief, while surgery and/or RT can eradicate disease in patients with isolated lung metastases [125]. As an example, one study of 26 patients with local recurrence showed that the survival at five years post local recurrence was 28 percent. Better survival outcomes were seen in those who did not have metastases at diagnosis of the recurrence, had a surgical treatment for the recurrence, and had complete eradication of all disease [145]. For patients with pulmonary recurrence who have not previously had whole-lung radiation, use of whole-lung radiation as a component of relapse therapy may prolong progression-free survival [155].

POSTTREATMENT SURVEILLANCE

Surveillance — There are no prospective data that address the appropriate schedule and type of surveillance for patients with ES after initial treatment. For patients with localized disease, we obtain a physical examination, complete blood count, chest imaging (radiographs or computed tomography [CT]), and surveillance imaging of the primary site every three months for two years, every six months for years 3 to 5, and annually thereafter. This is a similar approach to consensus-based guidelines from the National Comprehensive Cancer Network and from the Children's Oncology Group (table 3) [156-159]. For patients who presented with metastatic disease the frequency of imaging will depend on their treatment status and time since diagnosis. Patients who are on active treatment will need regular interval imaging at least every three months.

Local imaging of the primary site depends upon the specific site and prior local therapy. For example:

●Patients with a metal endoprosthesis are usually imaged with plain radiographs. The metal limits the utility of magnetic resonance imaging (MRI) in this setting, although metal subtraction techniques have improved the ability to assess the surrounding soft tissues in some instances. In addition, not all extendible endoprostheses are compatible with MRI due to the magnetic mechanism that allows for extension.

●Ultrasound may be useful to assess for soft tissue masses around an endoprosthesis or allograft with metal plate fixation.

●For patients treated with definitive radiation therapy (RT), MRI may be appropriate.

●For patients with chest wall primary tumors, a chest CT scan or chest radiograph may be used, depending upon level of concern for risk of recurrence.

Patients with symptoms or abnormal imaging or those who are planning surgical interventions may also be offered fluorodeoxyglucose (FDG) positron emission tomography (PET)-CT. A meta-analysis study showed that FDG-PET and PET-CT have a high accuracy in detecting distant metastases and postoperative recurrences in patients with ES [160].

We and others suggest that patients be followed indefinitely for the development of treatment-related toxicities and relapse [161]. Although most recurrences are observed within 10 years, patients with ES are at risk for late relapse and late development of treatment-related complications such as second neoplasms.

Clinicians performing posttreatment surveillance must be aware of concerns for radiation exposure and the risk for secondary malignancies, particularly in younger individuals. (See 'Complications in long-term survivors' below and "Radiation-related risks of imaging".)

Patients on surveillance who develop symptoms — Recurrent disease may be suspected in patients with symptoms at the primary site or elsewhere. Evaluation of suspected recurrence should include an assessment for disease at the primary site and for the presence of metastatic disease to guide therapy. Most patients with a local recurrence have either gross or microscopic metastatic disease. The evaluation for metastatic disease is similar to that done at the time of initial staging, as described separately. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma", section on 'Staging evaluation'.)

The documentation of a local recurrence can be challenging, especially in patients with metallic implants given distortion on imaging. In patients with metallic endoprostheses, MRI and CT evaluation can be distorted by metal artifact, although this is improved with contemporary MRI and CT imaging technology. In patients who are treated with certain extendable "growing" prostheses, MRI may be contraindicated. The interpretation of irradiated areas on imaging studies can be challenging because of the changes in bone caused by the radiation. Soft tissue masses may represent residual fibrosis rather than recurrent tumor. The evaluation of intraosseous sites is even more difficult since the response to prior therapy and the variability in remodeling may complicate the interpretation of radiographic studies. Progressive cortical destruction or increasing radiolucent areas suggest local recurrence, as do bone scans that demonstrate increased radiotracer uptake.

In patients with suspicion for recurrence we use the following guidance:

●PET scans may be useful to assess the likelihood of recurrence at a site that is suspicious on cross-sectional radiographic imaging.

●Ultrasound can also be useful to assess for soft tissue masses around an endoprosthesis.

●The decision to biopsy a site of suspected recurrent local or metastatic disease depends upon patient history and the level of evidence from imaging studies. Core needle biopsies can usually distinguish between recurrence versus reactive tissue/fibrosis and is the preferred initial biopsy technique. If this is nondiagnostic, then an open bone biopsy should be performed. However, open biopsies can be associated with local morbidity (ie, wound complications and bone fracture).

COMPLICATIONS IN LONG-TERM SURVIVORS — 

Long-term survivors of ES may have considerable burden of the late effects of their therapy and need to be followed lifelong for development of toxicities related to treatment [10,42,161-165]. Complications of chemotherapy include subsequent primary cancers, reduced fertility, kidney impairment, and cardiomyopathy. Complications of radiation therapy (RT) include subsequent primary cancers, pathologic fractures, wound complications, pulmonary fibrosis, neuropathy, limb leg discrepancy, and femoral head necrosis. The frequency and severity of these complications with modern treatment approaches are difficult to predict given the evolution of therapy and lag in data on long-term follow-up. Long-term follow-up guidelines after treatment of childhood malignancy are available from the Children's Oncology Group.

Secondary myelodysplasia (MDS) and leukemia are particular concerns for this population [166-170]. This was illustrated in a report from the Children's Oncology Group of 578 children with ES who were treated with three different regimens over a six-year period [168]. Overall, 11 children developed secondary MDS/acute myeloid leukemia, and the cumulative risk was higher among children treated with a regimen incorporating higher doses of doxorubicin, cyclophosphamide, and ifosfamide as compared with those receiving standard-dose vincristine, doxorubicin, cyclophosphamide, and dactinomycin with or without ifosfamide plus etoposide (11, 0.9, and 0.4 percent at five years, respectively).

The health status of long-term (≥5 years) survivors was addressed in a cohort study of 568 individuals who were diagnosed with ES before age 21 from 1970 to 1986, including a subset of 403 patients who were participating in the Childhood Cancer Survivor Study [164]. Cumulative mortality among all survivors was 25 percent at 25 years after diagnosis, and the cumulative incidence of secondary malignancy was 9 percent. Disease progression/recurrence accounted for 60 percent of all deaths, while other causes included secondary neoplasms, cardiac disease, pulmonary disease, and other medical causes. The cumulative mortality attributed to subsequent malignant neoplasms and cardiopulmonary disease potentially attributed to treatment was 8.3 percent at 25 years. In addition, compared with their siblings, survivors had higher rates of severe, disabling, or chronic health conditions; lower fertility rates; and higher rates of self-reported moderate to extreme adverse health status.

It is anticipated that the risk of late effects will decrease as a result of improvements in RT technique over the last 20 years (ie, the adoption of tailored radiation ports as opposed to whole-bone ports used in the period 1960 to 1980; three-dimensional conformal radiation therapy, intensity-modulated radiation therapy, and proton therapy; and use of lower and risk-adapted RT doses) [82,171]. In addition, trends in chemotherapy intensification may alter the pattern of secondary malignancies [163]. Nevertheless, the relatively high complication rate seen with earlier treatment approaches and the delayed nature of many of these complications underscore the need for long-term posttreatment follow-up.

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" and "Society guideline links: Bone sarcomas".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Ewing sarcoma (The Basics)" and "Patient education: Bone cancer (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Definition of localized disease – The treatment of Ewing sarcoma (ES) differs depending upon whether the patient has localized or metastatic disease. Localized ES is confined to the site where it originated (see 'Definition of localized disease' above). Metastatic disease refers to disease anywhere other than the primary site and can include involvement of lung, bone, bone marrow, distant lymph nodes, lesions discontinuous from the primary tumor (ie, skip metastases), or other distant sites.

●Multimodality therapy for localized disease – Patients with localized ES should be treated with multimodality therapy which includes neoadjuvant chemotherapy; local treatment such as surgery or radiation therapy (RT); and adjuvant chemotherapy (algorithm 1).

●Neoadjuvant chemotherapy – For patients with localized ES, we recommend initial chemotherapy with alternating cycles of vincristine/doxorubicin/cyclophosphamide and ifosfamide/etoposide (VDC/IE) rather than vincristine, ifosfamide, doxorubicin, and etoposide (Grade 1B), as this regimen improves overall survival (OS) and is shorter in duration with less toxicity. Neoadjuvant chemotherapy is given to reduce local tumor volume, facilitate resection, and treat micrometastatic disease. Patients receive 12 weeks of neoadjuvant therapy prior to local therapy. (See 'VDC/IE preferred' above.)

-Patients age <18 years – For patients age <18 years with localized ES, we recommend interval-compressed VDC/IE given every 14 days with hematopoietic growth factor support (table 1), rather than every 21 days (Grade 1B), as this approach improves event-free survival and OS without increased toxicity or higher risk of second (subsequent) malignant neoplasms. (See 'Interval-compressed VDC/IE' above.)

-Patients age ≥18 years – For patients age ≥18 years, we offer VDC/IE either every 14 or 21 days with hematopoietic growth factor support. (See 'Patients 18 years or older' above.)

•Local treatment – Either surgery or RT can provide effective local control. Tumor location, size, and impact of adjacent structures all influence choice of therapy. For tumors that can be resected with adequate margins and acceptable function and morbidity, we suggest surgery rather than RT (Grade 2C). RT is an acceptable alternative and preferred when the tumor is unresectable or if a resection will result in significant morbidity. (See 'Local treatment' above.)

When RT is delivered as local therapy, chemotherapy should not be interrupted. It should be delivered concurrently with RT (with omission of doxorubicin).

•Adjuvant Therapy

-Adjuvant chemotherapy – Adjuvant chemotherapy should be commenced as soon as possible after surgery and given concurrently with radiation for patients who do not have surgery. For patients receiving VDC/IE, we typically administer a total of 14 to 17 cycles which includes those cycles received as neoadjuvant therapy. (See 'Adjuvant chemotherapy' above.)

-Adjuvant radiation therapy – For patients with positive margins, gross residual disease, and/or high risk chest wall tumors, we suggest adjuvant RT (Grade 2C). Adjuvant RT is not usually offered to other populations. (See 'When to use adjuvant radiation therapy' above.)

●Metastatic disease

•Chemotherapy - Patients with metastatic ES at diagnosis receive multimodality treatment given with potentially curative intent as a subset of these patients may be cured. For most patients with metastatic ES, we suggest alternating cycles of VDC/IE (Grade 2C), based on extrapolation of data from trials in localized disease. However, since VDC/IE has not demonstrated a specific benefit for the addition of I/E in this population, some clinicians may opt for VDC for patients with unequivocal evidence of metastatic disease. (See 'Initial selection of chemotherapy' above.)

•Local therapy – In those with metastatic ES, local treatment of the primary tumor is performed as for patients with localized disease, though more of these patients are selected for definitive radiotherapy rather than surgery. In addition, local therapy of metastatic sites is offered to patients with isolated pulmonary metastases or solitary or circumscribed bone or soft tissue metastases. (See 'Local control in patients with metastatic disease' above.)

●Recurrent disease – Some patients with recurrent ES can achieve long-term disease control. The sites of recurrence, prior treatment, and relapse-free interval affect the treatment choices. Most patients with recurrent ES will receive systemic therapy prior to attempts at additional local control measures. (See 'Recurrent disease' above.)

●Posttreatment monitoring – Most relapses occur within two years of initial diagnosis, but later relapses have been observed. To monitor for relapse and late effects of therapy, patients are followed indefinitely with physical examination, bloodwork, and imaging (table 3). (See 'Surveillance' above.)

Lifelong follow-up is needed because disease relapse, treatment-related complications, and second malignancies can all occur beyond five years after treatment. (See 'Complications in long-term survivors' above.)

ACKNOWLEDGMENT — 

The UpToDate editorial staff acknowledges David C Harmon, MD, who contributed to earlier versions of this topic review.