Biological Sample Collection to Advance Research and Treatment: A Fight Osteosarcoma Through European Research and Euro Ewing Consortium Statement

Osteosarcoma and Ewing sarcoma are malignant bone tumors affecting ∼1,800 people annually in Europe (1). Despite continuous efforts and the investigation and intensification of treatment modalities, the prognosis for patients is poor when compared with other cancers (2, 3). Repeated attempts by large international cooperative groups to improve outcomes through randomized clinical trials have not led to survival improvement in osteosarcoma (4–11) and brought only modest benefits in Ewing sarcoma (12–18). A lack of available high-quality biological samples for omics (e.g., genome-wide profiling) assessments has meant that we still have poor understanding of the molecular basis of observed heterogeneous clinical phenotypes and mechanisms of chemoresistance and metastasis. Acquisition of snap-frozen and fresh tissue is recommended in international clinical guidelines (19–22) but is frequently not achieved and the absence of standardized procedures for sampling has hampered compliance.

Two international panels of scientists, clinicians, and patient and parent advocates formed the Fight Osteosarcoma Through European Research (FOSTER) consortium (www.fosterconsortium.org) and the Euro Ewing Consortium (EEC; https://www.ucl.ac.uk/cancer/research/centres-and-networks/euro-ewing-consortium/euro-ewing-consortium) to promote European collaboration and to accelerate clinical and scientific progress. The consortia have already delivered benefits by bringing together multiple—previously disparate—national clinical trial groups and scientists to develop and deliver collaborative trial protocols (14, 17, 23, 24), share samples (25, 26), and expertise (27) to perform collaborative research. A major goal of both consortia is the refinement and intensification of translational research. Systematic acquisition of high-quality biological samples from children and adults across multiple sites with associated clinical metadata should enable the identification and characterization of disease subgroups and tumor and germline genetic, biological, immunologic, and cellular environmental factors that can be used for the stratification of disease subgroup-specific therapies.

This position paper complements international clinical guidelines and provides comprehensive procedures for the adequate minimal sampling, handling, and storage of bone sarcoma samples that should be adopted across European centers. Although this statement has been drafted by the osteosarcoma and Ewing sarcoma communities, the principles discussed apply equally to other bone sarcoma histotypes and perhaps other cancers in which a lack of samples hinders translational research and clinical progress.

Key features of osteosarcoma biology include in utero loss-of-imprinting at chr.14q32 (28, 29), postnatal TP53 loss-of-function (or possibly mutant gain-of-function; refs. 30, 31) and complex genome rearrangements via chromoplexy and chromothripsis [(32, 33) and bioRxiv 2023.12.29.573403]. Specific molecular alterations in some cases include MYC amplification (34), RB1 deletion and mutation, and a “BRCAness” phenotype (35). Ewing sarcoma cells are characterized by gain-of-function gene rearrangements between FET (FUS, EWSR1, and TAF15) RNA binding proteins and ETS (FLI1, ERG, and FEV) transcription factors, most commonly EWSR1::FLI1 (36). The FET::ETS fusions encode oncogenic chimeric transcription factors with neomorphic features that reprogram the transcriptome (37), binding to GGAA microsatellites that become neoenhancers (38, 39), which leads to ectopic gene expression and tumor development. Additional STAG2 and TP53 cooperative mutations are associated with poorer survival (40–43).

Although the key driver mutations and recurrent alterations present in a subset of cases have been identified in both tumors, fragmented data from multiple small series and a lack of sufficient and appropriate solid and liquid tissue biopsies have hindered the development of molecular classifications and risk stratifications. Current and recent European clinical trials in Ewing sarcoma (ISRCTN92192408, ISRCTN36453794, and NCT00987636) have collected prospective liquid biopsies and accessed clinical diagnostic tissue samples to validate previously reported prognostic biomarkers, but none include specific molecular analysis of pretreatment and posttreatment tumor samples and clinical trials are not representative of all patient groups. For osteosarcoma, there have been no large prospective clinical trials since the closure of the EURAMOS-1 study and clinical trial samples do not inform individual patient treatment decisions. A culture of more universal prospective tissue collection is needed.

Preclinical models are a central component of translational research. Model systems such as patient-derived cell lines, ex vivo engineered models (44–46) and spheroids/tumoroids (47, 48), in addition to in vivo rodent (e.g., mice and rats), nonrodent (e.g., canine, zebrafish, and Xenopus; ref. 49), and chicken chorioallantoic membrane (50, 51) models, allow researchers to mimic bone sarcoma including its genetics and molecular biology, local microenvironment, systemic dissemination, and drug response. Most bone sarcoma deaths occur because of the emergence of drug-resistant lung, bone, and/or bone marrow metastases. Orthotopic and patient-derived xenograft (52) and engineered mouse models (53) recapitulating disseminated disease are essential. Sampling paired treatment-naïve and relapsed material is critical for the development of relevant models to avoid unfavorable scenarios in which preclinical drug efficacy data generated using less-relevant models seem promising (54–56), but the subsequent clinical trials show no patient benefit (23, 57, 58).

Historical cell lines, recent patient-derived cells, and orthotopic xenograft mouse models have been developed for osteosarcoma (52, 53, 59–63) and Ewing sarcoma (53, 56, 64, 65) but they typically over-represent the higher risk end of the disease spectrum. The Innovative Therapies for Children with Cancer consortium has generated patient-derived xenografts for in vivo compound testing from children with relapsed disease and includes some bone sarcoma models (66–68), but more representative and accessible patient-derived cell lines, xenograft, and genetically engineered autograft models that allow simultaneous examination of the tumor, immune, extracellular, and structural microenvironment are needed (69–72).

There is limited commercial incentive for the development of novel therapies for bone sarcoma. The European Medicines Agency (EMA) implemented the European Union Paediatric Regulation in 2006, requiring the investigation of new therapies in children before marketing authorization was granted. A waiver system in the initial legislation was modified in 2015 (73), strengthening the legal requirement to investigate all therapies with a relevant mechanism of action for childhood cancer. US Congress approval of the Research to Accelerate Cures and Equity for Children Act (“RACE Act”) enacted in 2020 gave the FDA powers to mandate pediatric clinical trials for new oncology drugs with a molecular target relevant to childhood cancers. There is considerable alignment between the EMA and FDA and this concerted regulatory approach has and will lead to greater opportunities for access to novel targeted therapies in children.

Although peaking in incidence in the second and third decades and occurring in older adults as well as children, osteosarcoma and Ewing sarcoma are frequently considered “pediatric” cancers and are represented in early phase pediatric drug trials. The regulatory coordination between the EMA and FDA presents an opportunity to utilize the inclusion of patients with osteosarcoma and Ewing sarcoma in early phase trials to study drug response and to develop predictive biomarkers. However, the number of patients with bone sarcomas recruited to each early phase trial is typically small (27, 74), sampling is not standardized, correlative biomarker studies are typically published long after conclusion of the trial, if at all, and together these factors have led to an extreme paucity of high-quality predictive biomarker evidence relevant to bone sarcoma. IGF1R inhibitors in Ewing sarcoma are an example of a failed opportunity to identify why only some patients responded to treatment. Across multiple early-phase trials, multiple agents, and more than 400 patients, IGF1R inhibitors resulted in response rates of 5% to 15%, including some sustained responses (75–80), but no predictive biomarkers were identified. As a result, no patient enrichment was possible in the Children’s Oncology Group AEWS1221 study comparing standard interval compressed VDC/IE with or without ganitumab. There was no significant difference in survival between the arms.

Several factors have converged to limit translational progress in osteosarcoma and Ewing sarcoma including recurrent molecular alterations not being validated, a consensus on molecular classification being made, a burgeoning of preclinical models but with an overemphasis on high-risk disease, a relative paucity of models for some disease settings, limited access to samples, almost nonexistent validated information about predictive biomarkers of response to cytotoxic chemotherapy, and molecularly directed treatment plus poor recruitment to early phase trials. In particular, while the key molecular drivers of osteosarcoma and Ewing sarcoma are relatively well understood, there is a fundamental lack of understanding of how genetic and epigenetic modifiers and tumor–host interactions affect disease progression and treatment response. This lack of understanding is largely driven by the absence of comprehensive, serial, annotated tumor tissue, normal tissue stroma, and liquid biopsy. At the level of clinical trials and collaborative large-scale research, there is a need for more, high-quality, tumor and normal tissue (solid and liquid) biopsies, ideally, serial biopsies to facilitate research into the molecular drivers and inhibitors of treatment response. At the level of individual patients, tissue acquisition needs to meet the needs of modern, multiomic analysis to monitor disease response and facilitate options for molecularly targeted, personalized medicine and critically for osteosarcoma, to identify patients with underlying cancer predisposition syndromes. Taking tyrosine kinase inhibitors (TKI) as an example, several TKIs have shown promise as single agents in osteosarcoma and Ewing sarcoma (81–86) but despite responses in up to 40% of patients, there are as yet no validated predicted biomarkers and the TKI mechanism of action remains obscure. Ongoing trials are evaluating combinations of TKIs with chemotherapy in first-line and relapse settings (e.g., the INTER-EWING-1 and rEECur trials developed by the EEC and NCT05691478 in the United States) and the FOSTER consortium was recently awarded ATTRACT funding to investigate the TKI cabozantinib as 12-month maintenance therapy following first-line standard therapy in osteosarcoma. All include sampling timepoints designed to investigate biomarkers predictive of TKI response.

A decades-long limitation to resolving some of the challenges discussed above is that there are no consistent or systematic Europe-wide practices for sample collection. Standard operating procedures (SOP) for biopsies and other sample types, storage, and sharing are either absent or only developed at local or national level. Exacerbating the problem is that there is little infrastructure and few dedicated staff to obtain bone sarcoma biopsies for both clinical care and translational research, although recent initiatives are working toward changing this landscape.

Across Europe, the stakeholders engaged in obtaining biopsy material have different practices. The amount, quality, and availability of viable tumor material is variable and frequently inadequate for molecular analyses. Because of the lack of a framework for sampling, much tumor tissue research is performed on postoperative, necrotic material obtained after induction therapy meaning there is “tainted” data and knowledge on tumorigenesis, clonal evolution, metastasis, and experimental drug response. There is evidence that the chromoplexy attribute of osteosarcoma results in dramatically different genetic alterations in different regions of the same tumor (bioRxiv 2023.12.29.573403), making a strategic approach to tissue biopsy critical to understanding patient-specific tumor biology and target actionability (Table 1).

We present consensus guidelines on the appropriate type and timing of tissue and liquid samples to facilitate research for future patients and to inform the treatment and future surveillance of current patients. Where this dedicated approach has taken place in other cancers, for example melanoma, the 10-year survival rate has improved from ∼10% (87) to ∼56% (88) because high-quality samples are made available for routine testing of the BRAF gene, which dictates first-line immunotherapy decision. Cytotoxic chemotherapy is now disregarded as first-line therapy in melanoma.

Cooperative effort from all involved disciplines is required. Routinely obtained written informed consent, collection, and storage of patient material for advanced biological studies is recommended in international clinical guidelines (19, 20, 22) but noncompliance exists because of the lack of standard procedures for biological sampling. Our position, complementing the clinical guidelines, is that all patients with bone sarcoma should have snap-frozen and fresh tissue samples (in addition to the conventional diagnostic samples) taken at diagnosis, surgery, and relapse regardless of their inclusion in research initiatives or clinical trials.

Biopsies should be performed at specialist bone sarcoma units (89, 90). Within research groups, clear definitions of the sample types and relevant SOPs should be used. Solutions for ethical, legal, and practical issues should be widely shared. To maximize the advantages of sample collection, to obtain a comprehensive biological understanding of bone sarcoma and host-related factors, different sample types at sequential stages of the clinical pathway should be collected (Fig. 1; Table 2). To enhance fundamental understanding of bone sarcoma clonal evolution and chemoresistance, tumor tissue collection at relapse and autopsy (e.g., PEACE study, NCT03004755) is essential. Metastases often comprise different genetics to the original primary tumor, so sampling metastatic lesions is recommended to ensure that the maximal amount of biological information is collected.

Treatment-naïve core or open biopsies should be obtained from suspected bone sarcoma cases at sarcoma specialist centers with the infrastructure to take, process and store (or send to a centralized national center) snap-frozen and fresh tissue in addition to the biological material placed in formalin. Fine-needle aspiration is not adequate. Biopsies and their position should be determined at a multidisciplinary team meeting with discussion on what the suspected lesion is expected to be, which tumor zones the biopsies should be taken from and by which approach to avoid unnecessary contamination. The procedure should be performed by a musculoskeletal or interventional radiologist experienced in the diagnosis of bone tumors or by a specialist surgeon and reported in line with the International Collaboration on Cancer Reporting (https://www.iccr-cancer.org/; ref. 20). The biopsy tract should be considered contaminated and resected en bloc during local therapy or be included in the radiotherapy field to minimize the risk of local recurrence (19, 21, 91–93). The surgeon who will perform the tumor resection should be involved in defining the optimal approach for the biopsy. The biopsy tract is preferably marked and described according to compartmental anatomy (91, 94, 95). In many cases, image-guided percutaneous biopsy using 8-, 11-, or 14-gauge needles represents a well-established alternative to open biopsy in terms of safety and diagnostic results (21, 91–93). Advantages and (contra)indications have been described for both procedures (91, 96–99).

Sampling focused purely on histologic diagnosis, usually from decalcified formalin-fixed, paraffin-embedded (FFPE) tissue, does not consider the developing prognostic technologies that require snap-frozen and/or fresh tissue that are becoming standards of care, for example, the macrophage expression phenotype in osteosarcoma (100). The equivalent of three 11- or 14-gauge needle biopsy samples have typically been sufficient to provide diagnostic yield (101) when paired with conventional histology. Our position is that where a core biopsy is performed, five samples should be collected where possible, of which at least one must be snap-frozen (Table 2). The fourth and fifth sample should be designated for research but can be used for diagnostic purposes in which a diagnosis could not be made using the FFPE samples. In many cases, the fourth and fifth sample will also be snap-frozen and stored but depending on active research studies, one or both could be formalin-fixed for use in spatial transcriptomics or used fresh for the isolation of live tumor cells for cell line generation, organoid development, and/or engraftment into immunocompromised animals (Table 2). For open biopsy, a minimum of 1 cm³ of tissue cut into multiple 0.2 cm³ sections is recommended. Where there are detectable oligometastases at presentation, consideration should be given to obtaining metastatic tissue at the time of the biopsy. For reference, recent Children’s Oncology Group guidance advocates up to 20 core biopsies for bone sarcomas with a soft tissue component (or up to seven core biopsies in which there is no soft tissue) plus up to three cores of underlying osteoid (102).

There are three surgical specimens in which resection serves as both performing standard of care and obtaining research samples: (i) primary tumor, (ii) matched adjacent normal tissue, and (iii) metastatic lesions. Samples should be prioritized by the pathologist to collect, depending on the availability of biobanking and specific research initiatives: (i) FFPE as the standard of care and neoadjuvant chemotherapy assessment, (ii) snap-frozen and stored, (iii) fresh and placed into an RNA-preserving medium, and (iv) fresh and placed into a culture-compatible medium (Table 2).

Samples from relapsed disease are particularly valuable if they can be paired with tissue from the primary diagnosis. As most bone sarcoma recurrences develop early and there is usually little doubt about the diagnosis, pretreatment biopsy material is scarcer than at initial diagnosis. Given the poor outcomes of relapsed disease and the limited treatment options, consideration should be given to obtaining snap-frozen and fresh and/or fixed tumor tissue at recurrence. These samples should be appropriately processed for omics assessment, other research, or biobanking. Irrespective of whether there are currently recruiting and/or routinely commissioned omics initiatives available at the time of recurrence, relapsed tissue is highly valuable if stored for future assessment.

Blood samples should be obtained at (i) diagnosis, (ii) before and after surgery, and (iii) at follow-up. Blood can be used as a liquid biopsy for the identification of ctDNA and circulating tumor RNA (ctRNA), circulating cell-free DNA, and circulating tumor cells (CTC). For specific diagnostic, monitoring and biomarker studies, urine and other body tissues (e.g., tears and hair) may be collected. Blood samples should be processed according to the relevant study, for example, CTC studies to be collected in cell-free Streck, PAXgene, or EDTA blood collection tubes (BCT) and processed immediately. Streck and PAXgene both have BCTs specifically designed for ctDNA and ctRNA capture. EDTA tubes can be used for either analytes, proteins, or live cells. There are pros and cons to each BCT related to the need for immediate versus delayed processing, plasma volume yield, and transport and storage costs. There is no consensus among European centers on which, if any, is best overall. We recommend that EDTA is used as a minimum for storage as these BCTs enable most analyses. But other more specific BCTs can be used according to research studies taking place at the time of collection. Blood samples may be key to detect micrometastases as well as allowing for the analysis of metastatic tumor-derived DNA, RNA (including microRNA), or proteins in circulation.

Technical aspects of collection and storage need to be considered to obtain minimum amounts of high-quality samples (Table 3), which may require a fundamental change in clinical practice in individual centers. Radiologists, surgeons, and pathologists have critical roles in the collection of adequate samples for histologic and molecular diagnostics and for translational research. The biopsy, operative, and histology procedures need to allow sufficient time to be devoted to sample collection and processing. These procedures should be appropriately funded. If diagnostic centers are unable to adequately process and store relevant material, consideration should be given by national bodies to restrict diagnostic biopsies to centers with adequate infrastructure or establish regulated delivery channels to central repositories.

SOPs for tissue processing should be implemented by designated staff other than the radiologist or surgeon because the tissue needs to be processed at the same time as the procedure being performed, which requires the full attention of the radiologist or surgeon. After collection, material allocated by the pathologist for diagnostic procedures will be processed as standard. Samples to be frozen should be transferred to sterile vials and immediately snap-frozen and stored in −80°C freezers or in liquid nitrogen. Fresh samples for cell and organoid cultures or animal engraftment need to be placed under sterile conditions into appropriate vials with a culture-compatible medium. The logistics and reagents may require preplanning with the research group for material transfer to the laboratory within 24 hours.

Sampling requires a team effort. Some centers will require changes to current care pathways, for example, automatic reminders to collect samples and duplicating processes so the biobank sample pathway is parallel with the pathology sample pathway. The radiologist’s and surgeon’s focus will be on the clinical procedure so it is important to establish a tissue processing pipeline as an interdisciplinary effort and adapt it to local conditions, which may include oncology, pathology, biobanking and theater staff. For SOPs to work, theater staff must be well informed, prepared, and adequately resourced to undertake the extra work. All personnel involved should recognize that tissue processing for research is pertinent to future patients being cured. Understanding the importance of their new role in tissue sampling could increase personnel efficiency and reliability.

Patients and their families overwhelmingly support research sample donation surplus to diagnostic requirement. FOSTER together with the Sarcoma Patient Advocacy Global Network has undertaken an international survey. The survey includes questions on diagnosis, treatment, and survivorship experiences, plus assessment of patient and family priorities for future research. Four questions are specific to sample donation. As of February 2, 2024, there were 372 combined osteosarcoma and Ewing sarcoma respondents (n = 234 osteosarcoma, median age 16 years; n = 138 Ewing sarcoma, median age 14 years). Just more than half of respondents with Ewing sarcoma (52.2%) and less than half with osteosarcoma (46.6%) were asked to donate research samples (Table 4). Of those asked, 97% consented to donate (Table 4). For the half of respondents who were not asked, almost two thirds reported that they would like to have been asked (Table 4).

Responding to patient-led direction involves important ethical, legal, privacy, and practical consideration (Table 5). Patients or their families must provide written informed consent for the collection, storage, and use of research samples. Lawful protocols should be in place to ensure that patient confidentiality and personally identifiable data are protected. Consent and protocols need to navigate the range of legal frameworks of different European nations. Age-appropriate information sheets for patients and their guardians must explain the purpose of the planned tissue storage and/or research, the recipients of the material (either now or in future), and the use of pseudonymized clinical data prior to providing forms for informed consent. Pairing sample data with pseudonymized clinical data including treatment and imaging findings and where the law allows, explicit linkage to regional and national cancer registries, should be possible. Ethical approval from international, national, or local authorities to study samples previously collected should be obtained.

Advantages of centralized versus decentralized (virtual) tumor banking and procedures to check for appropriate tissue representation for interpretable biological results should be considered. Whether sample availability should be defined as a mandatory inclusion criterion for patients going into clinical trials should be evaluated by regulatory bodies and ethics panels. For clinical trials, responsibilities of trial coordinators and local centers should be defined and adapted to applicable laws and regulations. Adequate coverage of the local costs and shipment of samples by research grants or national initiatives can help facilitate the compliance of local institutions, particularly where there are financial challenges faced by sample collection units. However, in some cases, financial constraints will prevent the collection of samples for unspecified research. Reusable tumor box devices can facilitate the shipment of frozen and unfrozen material. Practical aspects of exchange (including transborder) and the use of material should be defined by material transfer agreements (MTA) between research centers.

Biological material can be stored centrally by an academic tissue bank with software systems allowing for maximal up-to-date information about the stored materials. The materials can also be stored in local tumor banking facilities and later shipped in batches, as required, for use in further analyses. Both centralized and decentralized material storage allow for their use in big data analyses with bioinformatics support. Regardless of storage location (e.g., accredited laboratories with alarm monitoring vs. research lab freezers), proper evaluation by experienced bone sarcoma pathologists should ensure appropriate tissue representation before being used in specific projects. Biological material storage in aliquots allows for the tissue to be used for multiple research projects. Within existing legal frameworks of some European countries, it has been possible in some clinical units to store fresh and snap-frozen material from the biopsy before a diagnosis is obtained for a limited time prior to explicit patient consent for biobanking (103). This practice requires the appropriate infrastructure to be in place at the time of the biopsy and some bureaucracy to ensure adequate record keeping. Across most European centers, it is the tissue bank where the samples were collected that owns the biological material. Tissue banks are typically nonspecific repositories for all patient materials collected at a local institution or within a region or sometimes can be a study-specific biobank.

Transparent criteria for the regulation of access to larger material series by researchers from local contributing institutions could positively influence the cooperation of local centers. MTAs and SOPs for material shipment and adequate cost coverage (e.g., research grants) could further facilitate cooperative tumor banking. It is also important to establish procedures for the coupling of tumor material data to patient data. Genomic, transcriptomic, methylomic, and metabolomic data from tumor biopsies plus data from experiments on patient-derived cell cultures and xenografts should ideally be stored in an international bone sarcoma registry together with comprehensive anonymous clinical, radiological, and pathologic data. It is worth investing in the collection of large amounts of retrospective clinical data about baseline characteristics, treatment and survival from multiple international groups and to correlate these data with the analysis of genomic and epigenomic data from corresponding banked tumor samples. FOSTER, the EEC, and clinical trial groups should consider aspects of data collection and sample storage and discuss early in the planning phase of collaborative projects so that specific national requirements and future projects linking datasets can be implemented in a timely manner. Data sustainability beyond individual projects and connection of data at overarching levels should be considered.

Tangible progress in bone sarcoma has been bottlenecked by insufficient biological assessment and investigation, which in significant part has been caused by limitations in sample collection. Routine collection of decalcified and formalin-fixed tissue for histologic examination will not support diagnostic and prognostic technologies that evolve from translational research, for example, next-generation sequencing, in large part because fresh and snap-frozen tissue is not routinely stored. The benefits of obtaining fresh and snap-frozen samples at biopsy exceed the risks of complications of taking more tissue. Changing the process in which we collect biological samples and link patient data will lead to new molecular-based standards of care as well as new targeted therapies with fewer side effects.

Metastatic sites are not routinely sampled. Liquid biopsies are not routine. Screening for germline predisposition syndromes is not routine. The availability of properly sampled and stored biological materials will confer multiple scientific and clinical advantages including allowing identification and validation of new and reported prognostic factors and druggable targets. We need to ensure that children, teenagers, and young adults with bone sarcoma are not left behind while precision oncology offers new treatment solutions for more common, typically older adult, cancers. Because pediatric sarcomas are clinically and biologically highly distinct from adult cancers, precision medicine approaches should be adapted to make the best use of samples that are as informative as possible. Appropriate sample collection, storage, and sharing can only be achieved successfully if all the relevant steps are optimized at each local center. Collection and storage procedures could be adapted by local institutions to suit their individual structures, defined and assigned to dedicated individuals who are specifically educated and trained. FOSTER, EEC, and institutional researchers should actively collaborate, share data, methods, and samples, and disseminate good practice. These approaches will advance progress in bone sarcoma.

D. Andreou reports other support from PharmaMar outside the submitted work. S.S. Bielack reports personal fees from Y-mAbs, MAP BioPharma, SERB SAS, and Zschimmer and Schwarz Mohsdorf GmbH & Co. KG outside the submitted work. K. Boye reports personal fees from Bayer, GSK, Deciphera, Incyte, and NEC Oncoimmunity, as well as grants and personal fees from Eli Lilly and Company, personal fees from Novartis, and nonfinancial support from Merck outside the submitted work. N. Gaspar reports other support from ISPEN, Eisai, and Innovative Therapies for Children with Cancer outside the submitted work. J. Hardes reports grants from implantcast, Buxtehude, outside the submitted work. S. Hecker-Nolting reports grants from Deutsche Kinderkrebsstiftung, Bonn, Germany, and Foerderverein Krebskranke Kinder, Stuttgart, Germany, outside the submitted work. E. Palmerini reports personal fees from Deciphera Pharmaceutical, SynOX Therapeutics, Daiichi Sankyo, and Eusa Pharma outside the submitted work. S.J. Strauss reports personal fees from Ceridwen Oncology, Boehringer Ingelheim, Tessellate BIO, Inhibrx, and Adaptimmune outside the submitted work. C. Valverde reports grants and personal fees from PharmaMar, personal fees from Roche, GSK, Mundipharma, Eli Lilly and Company, Boehringer Ingelheim, and Ipsen, other support from Adaptimmune, Karyopharm, Foghorn Therapeutics, Ayala therapeutics, and Inhibrx, grants from Bayer, and nonfinancial support and other support from GEIS (Spanish Sarcoma Group) outside the submitted work. M.A.J. van de Sande reports grants from CarboFix and other support from SynOX and Deciphera outside the submitted work. M.G. McCabe reports other support from Eisai outside the submitted work. No disclosures were reported by the other authors.

This position paper was mainly written by D. Green, R. van Ewijk, E. Tirtei, Q. Campbell-Hewson, M.G. McCabe, and M. Nathrath. Data from the sarcoma experience survey were generated by P. Pantziarka. Authors are young investigators, work package and nation leads who all contributed to the content of the manuscript from the perspective of the group that they represent as well as commenting on general aspects of this policy review.

The Bone Cancer Research Trust (https://www.bcrt.org.uk/) funds FOSTER and EEC structure and activities. E.A. Roundhill and L. Cottone received funding from the Bone Cancer Research Trust (grant numbers BCRT/8422 and /7721). S. Hecker-Nolting received funding from Foerderkreis Krebskranke Kinder e.V. Stuttgart and the German Child Cancer Foundation (grant number DKS 2021.13). The consortia also thank Esperanza Perez, the Institut Gustave Roussy, Société Française des Cancers et leucémies de l’Enfant et de l’adolescent, ATTRACT, Enfants Cancers Santé, INTERSARC, une Main Vers l’Espoir, Imagine for Margo, The George Pantziarka TP53 Trust, German Cancer Aid (DKH), German Child Cancer Foundation (DKS), Society for Pediatric Oncology and Hematology (GPOH), Scandinavian Sarcoma Group, Trust Paola Gonzato Rete Sarcoma Onlus, Myrovlytis Trust, Ludovica’s Parente, InfoCilento, and The Christie Charity for more specific support.