Consensus recommendations regarding local and metastasis‐directed therapies in the management of relapsed/recurrent Ewing sarcoma

ABSTRACT

Limited randomized or prospective data are available to guide local/metastasis directed therapy (LMDT) in relapsed/recurrent Ewing sarcoma (RR‐ES), resulting in uncertainty regarding best clinical practice for these patients. This report reviews the available literature on LMDT approaches and provides consensus recommendations regarding therapeutic decision making, timing, and indications for the use of LMDT in the management of RR‐ES. LMDT should be considered on a case‐by‐case basis to assess appropriateness, optimal timing/modality, palliative versus curative intent, and its role in relation to chemotherapy. One commonly used LMDT is radiotherapy (RT), which can be delivered through standard, hypofractionated, or stereotactic techniques based on factors including prior RT, tumor size, and/or location. Chemotherapy can be combined with RT, although prospective data are limited in the relapse setting. Surgery for LMDT not only addresses the tumor but also provides tissue for analysis, though the potential surgical morbidity based on location, extent of resection, and recovery complications should be considered. Interventional radiology approaches also can procure tumor tissue while delivering LMDT; there are several different procedures available based on the location, size, and extent of disease. Finally, a combination of LMDT approaches can be used for patients with RR‐ES. Decisions regarding the management of RR‐ES should involve a multidisciplinary team and factor in the burden of disease, progression‐free interval, life expectancy, toxicity profiles of LMDT, and quality of life. In such patients, informed and shared decision making with patients and their families is paramount.

Short abstract

Limited data are available to guide decision making for local and metastasis‐directed therapy in patients with relapsed/refractory Ewing sarcoma. The current article reviews the available literature and provides consensus recommendations in support of therapeutic decision making, timing considerations, and indications to use various modalities in the management of relapsed/refractory Ewing sarcoma.

INTRODUCTION

Over the past several decades, the number of therapeutic options for patients with relapsed/recurrent Ewing sarcoma (RR‐ES) has significantly increased, leading to ambiguity in the management strategies for this patient population. ¹ It is clear that treatment of patients with relapsed disease should be geared toward maximizing outcomes, reducing toxicities, and supporting the patient's goals for care. Systemic chemotherapy is the backbone of RR‐ES treatment; however, specific metastatic lesions can cause or threaten symptoms and worsen quality of life. Therefore, they warrant local/metastasis–directed therapy (LMDT). Indeed, multimodal treatment is standard practice in the upfront management of Ewing sarcoma (ES), although clinicians may be hesitant to use local therapies because of potentially associated toxicities without clear survival benefit. Identifying how and when the use of LMDT may be of benefit for patients with RR‐ES is essential to optimizing the care of these patients.

Key questions to consider before the use of LMDT for RR‐ES include which modalities should be considered (e.g., radiation therapy [RT], surgery, and/or interventional radiology [IR]), in what clinical situations they should be considered, how to identify appropriate patients, and how best to sequence local therapy with systemic therapy approaches to maximize outcomes while minimizing overlapping toxicities. Currently, there are limited randomized or prospective studies to assist clinicians in managing patients who have RR‐ES with regard to LMDT. We previously published consensus recommendations for the systemic therapy management of RR‐ES, and the interest expressed in LMDT seemed substantial enough to warrant a dedicated report. ² Therefore, the purpose of this consensus statement is to review data on LMDT approaches, discuss appropriate utilization scenarios, and provide consensus recommendations on treatment approaches in the management of RR‐ES.

DISCUSSION

Question 1: Who should be offered LMDT?

When considering LMDT for RR‐ES, important factors to consider include time to recurrence, burden of disease, local symptoms, and prior therapy. For patients with more favorable characteristics, such as 2 or more years since initial diagnosis and a lower burden of metastatic disease, the incorporation of comprehensive local control may provide the most benefit because these patients have the longest expected survival. ³ However, most data for the combined/concurrent treatment of metastatic ES comes from the upfront/newly diagnosed setting. Recent clinical trials for RR‐ES incorporating local control are pending mature data, so care must be taken when extrapolating these principles to RR‐ES. Consequently, most nonpalliative local therapy for patients with RR‐ES commences after an initial response to salvage chemotherapy and may be especially helpful for sites that were slower to respond.

Local‐only recurrences are less common than distant relapses but are associated with improved outcomes. ³ , ⁴ Surgical management is preferred when negative margins can be achieved and the procedural risks are acceptable to the patient. If surgery is not performed or if margins are positive, RT should be considered. For previously irradiated sites, a treatment plan review is necessary to determine whether the recurrence is within the high‐dose radiation field, i.e., an infield recurrence, indicating radiation‐refractory disease, or outside of the radiation field. For out‐of‐field local recurrences, negative margin surgical resection, when feasible, is preferred over re‐radiation; however, RT may also be feasible for these lesions. In the setting of an infield recurrence not amenable to surgical resection in which standard fractionated RT was used upfront (1.8–2.0 grays [Gy] per fraction), the biologic effect of stereotactic body RT (SBRT) may result in improved local control. However, careful consideration is needed because there is a higher risk of toxicity with SBRT in a previously irradiated field. ⁵ Data for IR ablative treatments in RR‐ES are minimal; however, extrapolation from other sarcomas suggests that this therapy may also be used to increase the likelihood of local control for complex local recurrences (see question 5).

Given the dismal long‐term survival of patients with metastatic RR‐ES, which remains at approximately 10%–20% at 5 years, local therapy options should be evaluated cautiously, particularly for those with disseminated disease. ⁶ For patients with metastatic RR‐ES who have a favorable response to systemic therapy, LMDT can include surgical resection, SBRT or other definitive fractionated RT, and/or IR ablative treatments. In patients responding to systemic therapy, we recommend prioritizing LMDT to the site(s) of disease with the largest size and those with the most likelihood of impairing function or quality of life. All LMDT modalities should be considered; however, the treatment that provides the highest probability of local control with the lowest toxicity should be identified and prioritized. Disruptions in systemic therapy should also be minimized, so that LMDT can be interdigitated between cycles of chemotherapy after a disease response is confirmed. For patients who have progressive disease on systemic therapy, the benefit of local therapy needs to be considered carefully because these patients are less likely to benefit from aggressive management of specific lesions that are minimally symptomatic. LMDT given in this context, however, can still be helpful to relieve symptoms even when not expected to affect survival. For patients treated at institutions that do not have experience with advanced local therapy techniques, referral to tertiary institutions that offer these should be considered. Consensus statements are presented in Table 1.

TABLE 1.

Consensus statements.

Question
Who should be offered LMDT?	Any patient being treated with curative intent should be offered LMDT with RT, surgery, and/or IR ablative techniques, the choice of which depends on the specific clinical circumstances, risk/benefit, and patient preference. Also, LMDT may be helpful for those with symptomatic lesions seeking palliation.
When should RT be considered for LMDT?	RT should be considered for: (1)Recurrent primary tumors when surgery is not used or does not produce negative margins, (2)Consolidation of metastatic lesions responsive to chemotherapy, and (3)Palliation of symptomatic lesions.
How should systemic therapy be incorporated with RT as part of multimodality therapy?	Although systemic chemotherapy is the foundation for the treatment of RR‐ES, the addition of concurrent RT may offer a benefit over sequential RT in select cases. Factors like prior radiation exposure (e.g., WLI), radiation dose fractionation, available safety data of concurrent chemotherapy administration, patient performance status, life expectancy, and goals of care should all be considered. Treatment plans should be made on a case‐by‐case basis, with input from a multidisciplinary tumor board.
What is the role of surgical LMDT?	Surgery for locally recurrent tumors should be considered if negative margins can be achieved and the morbidity, functionality, and interruption to systemic therapy is acceptable. Surgery may also be useful in controlling some symptomatic lesions and can be considered for the management of metastases in select patients with limited tumor burden.
What IR approaches can be considered for LMDT?	Percutaneous ablation of liver and lung metastases can provide local tumor control in patients who are not surgical candidates, although data are less robust than data for SBRT. Cryoablation of osseous metastases is proven to be effective for pain management even in patients who failed RT. Preoperative embolization of Ewing sarcoma reduces blood loss during surgery.
How should CNS relapses be managed?	Patients with RR‐ES affecting the CNS have a poor prognosis, and the approach to therapeutic interventions should be individualized, come from expert multidisciplinary teams, and incorporate considerations of the disease burden, performance status, treatment burden/side effects, expected length of life, and the patient's goals of care.

Abbreviations: CNS, central nervous system; IR, interventional radiology; LMDT, local/metastasis–directed therapy; RR‐ES, relapsed/refractory Ewing sarcoma; RT, radiation therapy; SBRT, stereotactic body radiation therapy; WLI, whole lung irradiation.

Question 2: When should RT be considered for LMDT?

The role of stereotactic body radiation therapy

Because some patients who have limited metastases from sarcoma may achieve a cure with aggressive local treatment, SBRT may be an ideal approach because it delivers focused, high‐dose RT in one to five treatments, creating a noninvasive, locally ablative tool that spares normal tissues and increases the biologic effectiveness of RT. ⁵ , ⁶ , ⁷ Most existing data for SBRT in RR‐ES have been extrapolated from cohorts of patients with sarcoma who were treated for metastatic disease, including in the upfront setting. Brown et al. reported on 14 patients (27 lesions) who had recurrent or metastatic ES or osteosarcoma treated with SBRT, including one half who received treatment with curative intent. ⁸ Lesions were treated to a median dose of 40 Gy in five fractions, and the estimated local control rate at 2 years was 85% for lesions treated with curative intent. ⁸ A multi‐institutional phase 2 trial that examined the role of SBRT (40 Gy in five fractions) in patients who had sarcoma with bony metastatic sites reported the local lesion control rate was 95% with a median follow‐up of 6.8 months, with improvements in both overall survival (OS) and progression‐free survival (PFS) when all sites of osseous metastatic disease were treated with SBRT. ⁹ A phase 2 trial that evaluated SBRT for patients with metastatic or recurrent solid tumors, the majority of which were sarcoma, indicated that osseous targets rather than soft tissue had a higher rate of response according to the modified Response Evaluation Criteria for Solid Tumors. ¹⁰

Beyond outcomes, consensus guidelines for defining targets for nonspine osseous SBRT define clinical target volume (CTV) expansions for soft tissue extension or cortical disruption. ¹¹ Recent data suggest that a biologic effective dose (BED) cutoff of at least 95 Gy using an α/β ratio of 3 is associated with improved SBRT local control for more radioresponsive sarcomas, including ES. ¹² In addition, data and guidelines are available for normal tissue dose constraints to ensure safety. ⁵ , ⁸ , ¹³ Finally, SBRT should be considered cautiously for areas in which critical structure dosing may preclude safe treatment (e.g., abdomen, pelvis, head and neck, mediastinum/hilum).

The role of standard fractionation/moderate hypofractionation radiation therapy

Although SBRT is commonly considered the RT modality of choice for metastasis‐directed therapy, given the limited data for SBRT in the management of the primary tumors, there remains a role for standard RT (1.8–2.0 Gy per fraction) and moderately hypofractionated RT (2.5–3.0 Gy per fraction). ⁸ , ⁹ , ¹⁰ Standard and moderately hypofractionated approaches should be considered in patients who have inoperable primary tumors or for whom surgery is not recommended because of a significant impact on function or quality of life. In patients who have an uncontrolled primary tumor, standard fractionation can be considered with doses of 55.8–59.4 Gy in 31–33 fractions. Although standard fractionation has been well studied for the management of ES primary tumors, in the setting of metastatic disease, the timing of RT to the primary tumor remains controversial. Ideally, before pursuing long‐course therapy to the primary tumor, metastatic disease should be under control with a plan for definitive therapy. Moderate hypofractionation has been explored in the management of soft tissue sarcoma (STS) with promising results ¹⁴ ; extrapolating from this experience and using a dose of 55.8 Gy with standard fractionation, 15–20 fraction regimens can be considered, although limited clinical outcomes and toxicity data are available in the setting of RR‐ES. Hypofractionated RT may be necessary for larger tumors (i.e., >5 cm in length) or those close to critical structures where SBRT may not be safe. With regard to organs at risk when delivering moderate hypofractionation, hypofractionation guidelines from other diseases can be used for treatment planning. ¹⁵ , ¹⁶

Beyond treatment of the primary tumor, standard/moderate hypofractionation can be considered for larger metastases or the aforementioned anatomic areas that may preclude SBRT. In these situations, moderately hypofractionated approaches can be considered that allow for an elevated BED to provide appropriate tumor control while reducing the risk of acute and late toxicities. Hypofractionated RT can be combined with grid or lattice RT techniques, which are emerging as an option to overcome the radioresistance of larger tumors. ¹⁷ These therapies deliver radiation in a grid‐like or lattice pattern, creating high‐dose peaks surrounded by low‐dose valleys, which allows for a high BED (i.e., 15–20 Gy) delivered in a single fraction without major effects on surrounding critical structures.

Finally, as mentioned above, there may be instances in which RT is being considered solely for palliative intent. In these scenarios, lower total doses and fractions are used ranging from 8 Gy in a single fraction to 30 Gy in 10 fractions.

When not to consider radiation therapy

Most patterns of relapse in RR‐ES are not local but, rather, are distant, multifocal, and/or disseminated; these sites may best respond initially to systemic therapy approaches. ¹⁸ , ¹⁹ , ²⁰ Therefore, in many patients with RR‐ES, RT will only be considered to relieve symptoms even if it is not expected to affect survival. For patients with local failure, aggressive surgical management (e.g., amputation) may offer superior survival outcomes. ²¹ , ²² Of note, tumors that recur locally after chemoradiation may exhibit intrinsic radioresistance, limiting the utility of re‐irradiation approaches using conventional RT. In these patients, however, aggressive re‐irradiation delivery techniques, such as proton therapy or SBRT, which deliver a high BED for ablative intent, may have a role for localized relapses, as previously noted. ²³

Patients who experience intolerable, acute radiation‐related side effects or manifest severe late tissue toxicity from prior RT are generally poor candidates for additional RT. There is the potential for increased toxicity to regional tissues within the prior RT treatment field, exceeding its RT tolerance and, hence, the potential to cause significant injury, affect compartment function, decrease quality of life, and/or increase the risk of mortality. Although normal tissue dose tolerances are well characterized in the treatment‐naive setting, tolerances in the re‐irradiation setting are not clearly defined, and exceeding them can be associated with significant toxicity. ²⁴ , ²⁵ In addition, caution must be used when administering RT to local or distant recurrences in conjunction with high‐dose systemic therapy because there are many established augmented toxicities with the concurrent administration of chemotherapy and RT (see question 3).

Question 3: How should systemic therapy be incorporated with RT as part of multimodality therapy?

Primary systemic therapy augmentation with LMDT can be considered in certain situations, potentially offering additive benefit by leveraging different mechanisms of action. ²⁶ , ²⁷ In the case of treatment‐naive STS, sequencing permutations, including neoadjuvant, concurrent, adjuvant, and interdigitated (between cycles) approaches, have been reported. Unfortunately, there are limited data in the setting of RR‐ES, which means there is uncertainty about the increased potential for treatment‐related toxicity with combination therapy. ²⁸

Multiple systemic chemotherapy regimens exist for the treatment of RR‐ES (Table 2), some of which have established safety profiles for concurrent use with standard‐fractionation RT in the upfront management of ES (e.g., cyclophosphamide, ifosfamide, etoposide, vincristine). Other agents active in RR‐ES have established concurrent radiation safety profiles based on their use in other histologies (e.g., irinotecan, topotecan, temozolomide, trabectedin, gemcitabine, docetaxel), whereas others have known limiting toxicities (doxorubicin, actinomycin). ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ As newer systemic agents are used in the setting of RR‐ES, (e.g., multitargeted tyrosine kinase inhibitors), safety considerations for using concurrent or temporally proximal RT may need to be extrapolated from non‐ES–specific studies. For example, a Children's Oncology Group study (ClinicalTrials.gov identifier NCT02180867) used neoadjuvant chemoradiation plus pazopanib for STS and found no increase in toxicity with concurrent administration. ⁴⁴ There are prospective studies examining the tolerability and efficacy of concurrent administration of newer generation multitargeted tyrosine kinase inhibitors, such as cabozantinib, with radiation in treating sarcomas, and large multicenter data are suggestive of tolerability for concurrent radiation and multitargeted tyrosine kinase inhibitors in the treatment of renal cell carcinoma. ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ Combinations of RT concurrently delivered with topoisomerase I inhibitors or trabectedin have demonstrated adequate safety profiles in clinical trials and thus may be considered in future trials of RR‐ES. ⁴⁵ Importantly, the frequent use of hypofractionated radiation in the RR‐ES setting complicates extrapolation from standard‐fractionation–derived data. Even when safety profiles can be extrapolated, there is a need to carefully consider whether there is a known synergistic benefit or a high risk for locoregional complications. ⁴⁹

TABLE 2.

Chemotherapy regimens and utilization with radiotherapy.

Regimen	Agent and timing		Cycle length	Concurrent radiation	Adjuvant radiation	High‐grade treatment‐related toxicity
(V)IT (Raciborska 2013, 29 Xu 2023 30 )	Irinotecan 50 mg/m2 daily IV or	Days 1–5	21 days	No	Yes	No
	Irinotecan 90 mg/m2 daily orally
	Temozolomide 100–150 mg/m2 daily	Days 1–5
	(Vincristine 1.5 mg/m2 daily)	Day 1
(V)IT (Palmerini 2018, 31 Casey 2009 32 )	Irinotecan 10–20 mg/m2 daily IV or	Days 1–5, days 8–12	21 days	Yes	Yes	Yes, pneumonitis
	Irinotecan 35 mg/m2 daily orally
	Temozolomide 100–150 mg/m2 daily	Days 1–5
	(Vincristine 1.5 mg/m2 daily)	Days 1 and 8
TC (Hunold 2006 33 )	Topotecan 0.75 mg/m2 daily	Days 1–5	21 days	No	Yes	No
	Cyclophosphamide 250 mg/m2 daily	Days 1–5
HD ifosfamide (Ferrari 2009 34 )	Ifosfamide 3 g/m2 daily IV	Days 1–5	21 days	No	Yes	Ni
	Mesna 3 g/m2 daily IV	Days 1–5
GD (McCabe 2019 35 )	Gemcitabine 675–900 mg/m2 daily	Days 1 and 8	21 days	NR	NR	NR
	Docetaxel 75–80 mg/m2 daily	Day 8
	Dexamethasone 3 mg/m2 daily (up to 8 mg)	Days 7–9
ICE (Van Winkle 2005 36 )	Ifosfamide 1.8 g/m2 daily	Days 1–5	21 days	No	NR	No
	Carboplatin 400 mg/m2 daily	Days 1 and 2
	Etoposide 100 mg/m2 daily	Days 1–5
CE (van Maldegem 2015 37 )	Carboplatin AUC 2.2 daily or	Days 1, 8, and 15	21 days	NR	NR	NR
	Carboplatin AUC 7.5 daily	Day 1
	Etoposide 100 mg/m2 daily	Days 1–3
Trabectedin/irinotecan (Baldini 2018 38 )	Trabectedin 1 mg/m2 IV over 1 hour	Day 1	21 days	No	No	NR
	Irinotecan 25 mg/m2	Days 2 and 4
Cabozantinib (Raciborska 2013 39 )	Cabozantinib 60 mg daily (or 40 mg/m2 daily in patients younger than 16 years)	Days 1–28	28 days	NR	NR	NR
Regorafenib (Duffaud 2023, 40 Attia 2023 41 )	Regorafenib 160 mg daily (or 82 mg/m2 daily in patients younger than 16 years)	Days 1–21	28 days	NR	NR	NR
Oral etoposide (Kostos 2024, 42 Podda 2016 43 )	Oral etoposide 40–50 mg/m2 daily	Days 1–21	28 days	Yes	Yes	Yes, SMN

Abbreviations: AUC, area under the curve; CE, carboplatin and etoposide; GD, gemcitabine and docetaxel; HD, high dose; ICE, ifosfamide, carboplatin, and etoposide; IT, intrathecally; IV, intravenously; NR, not reported, SNM, second malignant neoplasm; TC, topotecan and cyclophosphamide; (V)IT, (vincristine), irinotecan, and temozolomide.

Pulmonary metastases are the most frequent sites of distant failure in ES, and de novo whole lung irradiation (WLI) for patients who have an isolated pulmonary relapse has demonstrated improved local control and PFS. ⁵⁰ In addition, WLI can include an integrated SBRT boost to residual masses after systemic therapy. However, oligometastatic progression of a few lesions in the setting of a wider response to systemic chemotherapy may justify targeted radiation of those lesions instead of WLI. Importantly, toxicities, including pneumonitis and restrictive fibrosis, may be increased with ablative chemotherapy regimens that use high‐dose chemotherapy, and cardiovascular toxicity from both alkylator chemotherapy and thoracic‐directed RT can be significant. ⁵¹ , ⁵² , ⁵³ Therefore, general practice patterns dictate holding systemic chemotherapy when patients receive lung irradiation/SBRT to oligometastatic lesions, and this is unlikely to compromise the systemic control achieved with the chemotherapy regimen if the radiation course is short.

The use of concurrent chemoradiation for osseous, soft tissue, and lymph node metastases in RR‐ES has not been established. In the upfront setting, standard practice has evolved to include radiation of a distant metastatic site after the completion of chemotherapy, for example, Children's Oncology Group studies AWES1221 and ARST2031 (ClinicalTrials.gov identifiers NCT02306161 and NCT04994132, respectively). If using RT in the relapsed situation, similar consideration can be given. ⁵⁴ , ⁵⁵

Question 4: What is the role of surgical LMDT?

Surgical considerations in the setting of RR‐ES can be separated into local and metastatic sites of disease and must be viewed in the context of prognostic factors. One analysis demonstrated a 5‐year PFS of 31.4% for patients with RR‐ES who underwent resection versus 9.1% for those who did not undergo surgery (p = .02), with studies confirming that those who had recurrences limited to the primary tumor or the lung had better outcomes. ²¹ , ²² These data must be interpreted in the context of selection bias, namely, whether the ability to undergo surgery or the surgery itself portends a better prognosis.

For local recurrences, discussion in a multidisciplinary tumor board to ensure consideration of relevant patient variables (see question 1) is paramount. Appropriate surgical consultation (orthopedic oncology, pediatric surgery, thoracic surgery, general surgery, etc.) allows for assessment of the probability of marginal re‐resection, the effect of surgery on overall disease prognosis, the timing of the intervention, and the effect on function and quality of life. The goal for surgery should be negative margins. If pathology review indicates that the margins are positive, adjuvant RT should be considered; additional considerations for RT can be based on the response observed at the time of surgery if preceded by systemic therapy. In appendicular relapses, especially at the site of prior definitive RT, surgical resection, including amputation, may be indicated if it will render the patient disease‐free, or for the treatment of intractable pain or a fungating tumor. Additional considerations include patients who have a better prognosis in the setting of a localized relapse, especially if wide margins cannot reliably be achieved with limb‐salvage resection. However, in multifocal osseous, soft tissue, and lung relapses, radiation to the relapsed primary site over an amputation should be considered. It is also important to consider the location of the relapse in the limb because recovery from a distal amputation may be better tolerated compared with more proximal sites. In the setting of bony resection, reconstruction options should allow for early mobility and recovery to enable patients to resume systemic therapies. Some axial recurrences may not be amenable to resection, and radiation may be the only option.

Modern outcomes in pediatric patients with ES of the chest wall demonstrate 78%–88% local 5‐year recurrence‐free survival after initial combined treatment approaches and improved outcomes with an R0 (complete) resection; of the one quarter of patients who had a local recurrence, one half died despite salvage therapy. ⁵⁶ For chest wall local recurrences, surgical resection may be feasible even when there is a history of prior surgical resection and reconstruction. Sternal involvement and chest wall resections involving three or more ribs require rigid reconstruction (for example, mesh with methyl methacrylate) to optimize chest wall mechanics and to protect underlying structures. Near the scapular tip, even smaller defects may require reconstruction to prevent scapular entrapment. For chest wall tumors involving the diaphragm or pericardium, these structures can be resected en bloc and reconstructed as well. Tumors involving the lung may also require en‐bloc lung resection. For tumors that mainly extend into the pleural space, hybrid thoracoscopic approaches may minimize the risk for positive margins or disrupting the tumor. Chest wall tumors at the superior sulcus are difficult to expose, and these may require attention to vascular structures (anteriorly) and the brachial plexus and/or spine, potentially requiring multidisciplinary surgical expertise in the operating room. Prior chest wall reconstructions may be spared at re‐operation. If the local recurrence occurs more than 3 months after the initial operation, removal of a prior mesh reconstruction may be acceptable (as long as the area does not require rigid support) because a fibrous layer usually forms under the mesh reconstruction over time that can provide adequate support if left intact.

With regard to metastatic lesions, surgery for pulmonary relapse is a controversial topic, and a family oriented multidisciplinary decision‐making process is recommended. Given both the significant practice pattern bias in the retrospective literature and the lack of data on both beneficial and adverse outcomes necessary to support prospective evaluation, it is difficult to discern whether surgery offers a survival advantage in the setting of pulmonary relapse. In contrast to osteosarcoma, this benefit has not been proven in RR‐ES, in which relapse is considered more of a systemic disease for which chemotherapy is used. As mentioned above, the possibility exists that those who are candidates for surgical resection simply have favorable underlying tumor biology. ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ There is a growing body of evidence to support the use of tumor biology—or surrogate markers of underlying biology—to guide clinical decision making. The latter scenario likely plays a significant role in reports describing improved survival after resection of pulmonary relapse as well as recent data on outcomes for those with a rapid complete response. ⁶¹ A 30‐year cohort consisting of patients who had multiple primary pathologies (osteosarcoma, chondrosarcoma, and ES) demonstrated an overall postmetastasectomy 5‐year survival rate of 43%, with results suggesting that a disease‐free interval >18 months, greatest metastasis diameter <1.8 cm, and complete surgical resection were associated with improved outcomes. ⁶² A review of 71 patients treated from 1979 to 1999 supports the notion that a longer disease‐free interval and isolated relapse (local or pulmonary as opposed to both) are relatively good prognostic indicators for patients with RR‐ES; another case series of 26 mixed primary malignancies corroborated the need to achieve complete surgical clearance of pulmonary relapse, if attempting resection, to optimize outcomes. ²¹ , ⁶³ Other prognostic indicators that may be incorporated into a decision‐making algorithm include the disease‐free interval, nodule size and number, and the anticipated ability to surgically clear all pulmonary disease.

Pulmonary resection may also be considered in cases of diagnostic uncertainty, such as for a solitary peripheral nodule with imaging suggestive of relapse or oligometastatic disease located in the periphery of the lung that is amenable to complete resection by nonanatomic resection (i.e., pulmonary wedge resections), although the benefit of this approach compared with systemic therapy or WLI/SBRT is unclear. Central or hilar pulmonary relapse typically requires anatomic lung resection (i.e., lobectomy). In this situation, surgery should be limited to select patients with an isolated pulmonary relapse and a relatively long disease‐free interval, acknowledging that evidence supporting this practice may be affected by selection bias. ⁶⁴ Pneumonectomy is generally not considered advisable for relapsed ES, although each patient should be considered on a case‐by‐case basis. To maximize lung preservation, precision cautery is a parenchyma‐sparing surgical technique that involves using a pinpoint cautery tip to carve out a lung nodule with a thin margin of surrounding lung tissue. The resulting wound bed is re‐approximated using absorbable suture. Care must be taken not to simply close the pleural surfaces at the cut edge, as this leads to the formation of cysts. This approach can be used when lung metastases are too numerous to resect in any other fashion or as an option to avoid a larger anatomic resection for central lesions that do not involve major vessels or airways.

Considering the differing advantages and disadvantages of each local approach, combinations of therapies may also be considered when the goal is to direct local treatment to multiple lesions in a parenchyma‐sparing manner. For example, surgical resection may be used to resect multiple peripheral lesions with parenchymal‐sparing resections followed by RT for a single central lesion that otherwise would have required an anatomic resection or ablation for lesions that are too small to be identified with surgery but can be localized with an image‐guided therapy.

Question 5: What IR approaches can be considered for LMDT?

IR has emerged as the fourth pillar of cancer care alongside medical oncology, surgical oncology, and radiation oncology and has played an evolving role in the management of sarcomas. IR applications have ranged from image‐guided biopsies of suspected recurrent tumors and the implantation of vascular access devices to interventions for cancer‐related pain and as well as serving as a primary treatment modality. ⁶⁵ Multiple therapeutic IR techniques are available, including percutaneous and transarterial embolization approaches, as well as embolization approaches before surgery to reduce blood loss. ⁶⁶ , ⁶⁷ Unfortunately, outside of case reports, there is a lack of primary literature supporting the role of IR therapies in treating ES, although data from experiences of applying IR approaches in other sarcomas can be considered.

Percutaneous ablation offers an alternative to surgical resection for smaller tumors (<3 cm) situated in locations amenable to ablation and for patients who are not suitable candidates for surgery. Multiple approaches exist, including radiofrequency ablation, microwave ablation (MWA), high‐intensity focused ultrasound, and cryoablation (CA), and studies have evaluated the role of these approaches in the management of lung, osseous, and liver lesions. ⁶⁸ With regard to lung metastases, an analysis of 566 patients (51 with sarcoma) with 1037 metastatic lesions of the lung who were treated with computed tomography–guided percutaneous thermal ablation reported 1‐year, 3‐year, and 5‐year treatment failure rates of 6.1%, 8.3%, and 8.3%, respectively; mean PFS rates of 43%, 26.5%, and 15.9%, respectively; and mean OS rates of 94.1%, 58%, and 41.5%, respectively for the sarcoma cohort, with similar outcomes observed in additional series. ⁶⁹ , ⁷⁰ , ⁷¹ , ⁷² Overall, it has been demonstrated that thermal ablation is safe and effective in the setting of sarcoma oligometastatic to the lung; however, limited data are available for comparing these approaches with surgery and SBRT, which are considered the current standard. In addition, evidence for percutaneous ablation has demonstrated efficacy for pain control and locoregional disease control in osseous tumors. Fan et al. reported outcomes of MWA in 104 patients with primary malignant pelvic bone tumors, including 11 patients with ES. ⁷³ The patients who had ES had a 91% 3‐year mean OS rate after MWA. Common Terminology Criteria for Adverse Events (CTCAE) grade 3 or greater complications were observed in 18% of all patients and included infection, fistula, lumbosacral trunk damage, sciatic notch fractures, and a death. An alternative to MWA is CA, which allows for controlled tissue destruction while monitoring the formation of an ice ball during the procedure, thereby minimizing the risk of damage to adjacent structures. The MOTION study (ClinicalTrials.gov identifier NCT02511678) was a multicenter, prospective study that demonstrated the effectiveness of single‐session CA to treat moderately to severely painful osseous metastases in 66 patients who were either ineligible for or unresponsive to standard treatments. ⁷⁴ In that study, pain improved as early as 1 week after CA, with clinically meaningful improvements observed on average starting from week 8 onward. Adverse events possibly related to CA occurred in 22% of participants, including hematoma, nausea/vomiting, tumor or needle pain, hypotension, pleural effusion, skin burn, and frostbite, and patients had grade 3 or 4 events, including abdominal pain, hematoma, and a grade 4 skin burn that led to a below‐the‐knee amputation. Similar to pulmonary ablation techniques, limited data are available for definitive local therapy or comparing these approaches versus traditional treatment approaches for osseous metastases. With regard to liver metastases, outcomes with percutaneous ablation in treating liver metastases of ES have not been reported, possibly because of their rarity; however, experience with other sarcomas oligometastatic to the liver has demonstrated the safety and efficacy of this approach, although rare complications included mild cytopenias and pleural effusion in approximately 5% of patients have been reported, with death from cryoshock multiorgan failure in 1% of patients. ⁷⁵ Finally, with regard to primary tumors, some preliminary data are suggestive of the feasibility of high‐intensity focused ultrasound in conjunction with systemic therapy, with minor adverse side effects, such as local pain and edema; skin toxicity in 21%, including a patient with a third‐degree burn; peripheral nerve damage in 12%; bone fracture in 8%; and lesser rates of ligamentous laxity, epiphysiolysis, or secondary infection. Further studies are needed at this time. ⁷⁶

Transarterial embolization, transarterial chemoembolization (TACE), and transarterial radioembolization (TARE) have shown promise in STS treatment, including symptom management and prior to resection. With regard to pain control, a series of 10 patients with STS who underwent conventional lipiodol TACE using doxorubicin or cisplatin followed by gelfoam embolization demonstrated that eight of nine patients (89%) who had pain experienced a substantial decrease in pain after embolization, and eight of 10 patients (80%) achieved significant devascularization of their tumors. All patients experienced transient fever and pain, which resolved within 1 week. There was one reported case of gluteal necrosis, and no additional CTCAE grade 3 or 4 toxicities with similar outcomes were reported in additional series. ⁷⁷ , ⁷⁸ The role of embolization before surgery has also been evaluated and may be considered for patients undergoing surgical resection for a metastatic lesion or recurrent primary. A retrospective analysis of 31 patients who underwent preoperative transarterial embolization of hypervascular osseous and soft tissue tumors, including metastases, demonstrated significant devascularization of their tumors, and all patients had complete or near complete devascularization. Intraoperatively, 29% of patients required minimal transfusion (median, 3 units of packed red blood cells). ⁷⁹ With regard to control outcomes after transarterial therapies, the majority of data come from liver‐directed therapies (LDTs), including data for the combination with systemic therapies. ⁸⁰ Krzyston and colleagues examined LDTs in 24 patients with metastatic leiomyosarcoma to the liver (n = 13). Interventions included DEB‐TACE (TACE with drug‐eluting beads; n = 13), TARE (n = 6), ablation (n = 5), and a combination of LDT (n = 3). CTCAE grade 3 or 4 toxicities were seen in three of 24 patients (12.5%). The median PFS was reported to be 9 months. DEB‐TACE had lower rates of progression (13%) compared with TARE (50%) and ablation (60%). ⁸¹ Other groups have studied the combination of yttrium‐90 TARE with systemic therapies.

Question 6: How should central nervous system relapses be managed?

RR‐ES affecting the central nervous system (CNS) is a rare and prognostically poor event that poses local control challenges, with median event‐free survival and OS after CNS disease detection of 1.9 months (range, 0.4–10.3 months) and 4.6 months (range, 1.1–50.9 months), respectively. ⁸² There is no standard definition for CNS involvement, but this can include disease in the brain parenchyma, leptomeninges, dura mater, and/or cerebral spinal fluid. Consensus treatment approaches are lacking for ES metastatic to the CNS, and the low incidence (3%–10%) makes prospective studies infeasible. ⁸³ , ⁸⁴ Case reports/series document clinical practice variability; approaches include focal radiation, craniospinal radiation, surgical resection, and IT chemotherapy with or without a systemic chemotherapy backbone. ⁸³ , ⁸⁴

RT is a well described local therapy for solid tumors metastatic to the CNS and can be used for the treatment of ES metastatic to the brain parenchyma. For amenable lesions, treatment with standard or hypofractionated RT is common, targeting a BED of 50 Gy. Radiosurgery can also be considered for patients who have small/unresectable parenchymal lesions. The role for craniospinal radiation is not established, but its use has been reported in patients with diffuse leptomeningeal disease and multiple foci of CNS disease (median dose, 36 Gy), although some authors recommend craniospinal radiation for any CNS involvement of ES. ⁸⁴ Surgical resection can also be considered in the setting of limited brain parenchymal metastasis. Outside of emergent neurosurgical intervention for symptomatic lesions, the value of resection in the relapsed setting is not clear. Given the low likelihood of a cure in the relapsed setting, a multidisciplinary discussion that includes consideration of the burden of disease outside of the CNS, location and resectability of the lesion, the possibility of morbidity after surgery, and patient performance status should inform any surgical plan. The use of IT chemotherapy (topotecan, etoposide, or methotrexate) has been described in combination with systemic chemotherapy and/or surgery. ⁸⁵ Of note, in the newly diagnosed setting, prophylactic CNS radiation and IT methotrexate did not decrease the incidence of CNS recurrence. ⁸⁶ Although no efficacy data exist to support the use of IT chemotherapy in ES, National Comprehensive Cancer Network guidelines for leptomeningeal metastases of solid tumors include a recommendation for intracerebrospinal fluid chemotherapy as a treatment modality in select patients. ⁸⁷

CONCLUSIONS

The consideration of LMDT for patients with RR‐ES requires multidisciplinary discussions along with informed and shared decision making with patients and their families. Patients treated with either curative or palliative intent should be evaluated for LMDT, although the considerations and decisions may be different for each group. Decisions about when to implement therapy based on the extent of disease, progression‐free interval, life expectancy, and goals of care with consideration of the toxicity of local therapies are paramount given the unclear effect on survival for this cohort of patients. Future directions include the need for prospective trials to evaluate the optimal timing for application of LMDT and how to effectively couple this approach with novel systemic therapies.

AUTHOR CONTRIBUTIONS

Chirag Shah: Conceptualization, writing–original draft, writing–review and editing, and supervision. Shauna R. Campbell: Conceptualization, writing–original draft, and writing–review and editing. Erin Murphy: Conceptualization, writing–original draft, and writing–review and editing. Steve Braunstein: Conceptualization, writing–original draft, and writing–review and editing. Matthew S. Dietz: Conceptualization, writing–original draft, and writing–review and editing. Odion Binitie: Conceptualization, writing–original draft, and writing–review and editing. Zachary J. Kastenberg: Conceptualization, writing–original draft, and writing–review and editing. Jane Yanagawa: Conceptualization, writing–original draft, and writing–review and editing. Jennifer Halpern: Conceptualization, writing–original draft, and writing–review and editing. Bela Kis: Conceptualization, writing–original draft, and writing–review and editing. Stephen Hunt: Conceptualization, writing–original draft, and writing–review and editing. Fereshteh Yazdanpanah: Conceptualization, writing–original draft, and writing–review and editing. Ajay Gupta: Conceptualization, writing–original draft, writing–review and editing, and supervision. Matteo Trucco: Conceptualization, writing–review and editing, and supervision.

CONFLICT OF INTEREST STATEMENT

Chirag Shah reports grants/contracts from PreludeDX and Varian Medical Systems Inc.; and personal/consulting fees from ImpediMed Inc., PreludeDX, and Videra Surgical outside submitted work. Erin Murphy reports personal/consulting fees from Autonomix outside the submitted work. Odion Binitie reports personal/consulting fees from Onkos Surgical Inc. outside the submitted work. Stephen Hunt reports personal/consulting fees from Boston Scientific Corporation, GE Healthcare, GEURBET LLC, and Varian Medical Systems Inc.; and stock ownership in Vivaldi Therapeutics outside the submitted work. The remaining authors disclosed no conflicts of interest.

Shah C, Campbell SR, Murphy E, et al. Consensus recommendations regarding local and metastasis‐directed therapies in the management of relapsed/recurrent Ewing sarcoma. Cancer. 2025;e35858. doi: 10.1002/cncr.35858