Sarcoma care in the era of precision medicine

Sarcoma subtype classification is currently mainly based upon histopathological morphology. Molecular analyses have emerged as an efficient addition to the diagnostic workup and sarcoma care. Knowledge about the sarcoma genome increases, and genetic events that can either support a histopathological diagnosis or suggest a differential diagnosis are identified, as well as novel therapeutic targets. In this review, we present diagnostic, therapeutic, and prognostic molecular markers that are, or might soon be, used clinically. For sarcoma diagnostics, there are specific fusions highly supportive or pathognomonic for a diagnostic entity—for instance, SYT::SSX in synovial sarcoma. Complex karyotypes also give diagnostic information—for example, supporting dedifferentiation rather than low‐grade central osteosarcoma or well‐differentiated liposarcoma when detected in combination with MDM2/CDK4 amplification. Molecular treatment predictive sarcoma markers are available for gastrointestinal stromal tumor (GIST) and locally aggressive benign mesenchymal tumors. The molecular prognostic markers for sarcomas in clinical practice are few. For solitary fibrous tumor, the type of NAB2::STAT6 fusion is associated with the outcome, and the KIT/PDGFRA pathogenic variant in GISTs can give prognostic information. With the exploding availability of sequencing technologies, it becomes increasingly important to understand the strengths and limitations of those methods and their context in sarcoma diagnostics. It is reasonable to believe that most sarcoma treatment centers will increase the use of massive‐parallel sequencing soon. We conclude that the context in which the genetic findings are interpreted is of importance, and the interpretation of genomic findings requires considering tumor histomorphology.


Etiology and incidence
Sarcomas encompass a diverse group of mesenchymal malignancies, which can be broadly categorized into soft-tissue sarcomas and malignant bone tumors [1].These tumors are divided into more than one hundred histological subtypes and occur in any organ [2].The incidence varies between countries and is not fully known, but it is estimated to be around 1/10,000/year [3].Approximately 1% of all malignancies in adults are classified as sarcomas, and up to 20% of childhood malignancies [1].
The etiology of sarcomas is largely unknown but differs from that of many other solid tumors.For instance, smoking has not been shown to have a significant impact on soft-tissue sarcoma incidence [4].Environmental factors may increase the risk of sarcoma, such as radiotherapy or extensive lymph node surgery in the axilla [1].It has been suggested that half of the sarcoma patients could have potentially pathogenic monogenic and polygenic variations in cancer genes [5].However, most sarcomas occur without family history and with no known association with a genetic cancer syndrome.There are cancer syndromes that include a known increased risk for sarcomas, such as heritable TP53-related cancer syndrome, familial adenomatous polyposis, Noonan syndrome, Rubinstein-Taybi, Beckwith-Wiedemann, and neurofibromatosis type 1 [6].

Clinical modalities for diagnostics
Soft-tissue sarcomas generally present as nodules, which may or may not be painful.Signs of malignancy include deep tumor localization, a dimension exceeding five centimeters, a recent increase in size, and pain [7].Malignant bone tumors often cause persistent nonmechanical pain lasting for weeks or months.If the tumor has progressed through the bone cortex and distended the periosteum, soft-tissue swelling may occur [8].The initial clinical workflow for suspected sarcomas includes imaging (magnetic resonance imaging for diagnosis and local staging, complemented with conventional X-rays for bone sarcomas), and other techniques such as computed tomography and positron emission tomography for staging of suspected distant metastases, and percutaneous needle biopsies for a histological diagnosis [9][10][11].Although most centers use core needle biopsies, fine-needle aspiration biopsies are also used in centers with cytopathology expertise.When the diagnosis is uncertain, surgical biopsies may be motivated.
The initial histological diagnosis of sarcoma has been shown to be inaccurate in 3%-14% of cases, and these inaccuracies may be corrected through expert panel review or additional auxiliary tests, such as genetic analyses [12][13][14].Prendergast et al. showed that whole-genome sequencing (WGS) findings prompted the modification of the initial histopathological diagnosis in 8/350 patients (3%), and the concordance between WGS and standard of care testing was in general high [13].Italiano et al. argued that "even when sarcoma cases are assessed by an expert pathologist, molecular genetics can provide relevant and independent information."In their approach, the initial histopathological diagnosis was altered in 14% of the cases after applying molecular tests that were not included in standard of care diagnostics [14].Accurate diagnoses are crucial for appropriate treatment and follow-up, as well as conducting research studies.There are numerous studies and quality registries showing improved patient outcomes when sarcoma diagnostics and treatment are centralized [9,15,16].Improved diagnostics can potentially lead to better overall survival and reduced morbidity [17].

Clinical pathology
The classification of mesenchymal tumors poses significant challenges in clinical pathology due to the rarity of individual subtypes and their indistinct histomorphological features.Soft tissue and bone tumors are classified using the widely accepted World Health Organization (WHO) Blue Books system, which undergoes regular review by expert pathologists, geneticists, and clinicians.Although the classification has evolved over the years, the main diagnostic groups delineated by histological lines of differentiation (such as adipocytic, fibroblastic/myofibroblastic, fibrohistiocytic, and vascular) remain largely intact [18].The precise cell of origin for mesenchymal tumors remains undisclosed.Some entities likely stem from their normal counterparts, such as many vascular lesions.Others, such as rhabdomyosarcomas, rarely arise in skeletal muscle and its differentiation reflects the underlying genetic reprograming of the cancer cell.It has been suggested that many sarcomas may originate from mesenchymal pluripotent stem cells [19,20].Others have suggested that sarcomas may stem from different cell types, and the cell of origin partly determines the histological phenotype [21].
Soft tissue and bone tumors are classified either as benign, locally aggressive, rarely metastasizing, or malignant, reflecting the width of their clinical behavior.Some high-grade sarcomas may have insidiously benign-looking cells and low proliferation, whereas some reactive or benign lesions may have cytological atypia and high proliferation, often associated with malignancy [18].When evaluating musculoskeletal tumors, it is almost always necessary to correlate the histological findings with radiology to reach an accurate diagnosis.

Laboratory techniques
The histopathological workflow for sarcoma diagnostics is mainly based on macroscopic and microscopic morphology, but immunohistochemical markers are commonly employed to support diagnostic evaluations.Additionally, various auxiliary methods can be utilized to provide evidence of underlying genetic alterations and assist in making accurate diagnoses [9].Moreover, bone tumors require special laboratory handling, including decalcification of tissue or sawing of large specimens [22].
A growing number of markers have been identified by differential gene expression studies, including MUC4 in low-grade fibromyxoid sarcomas [34] There are likely more biomarkers to be discovered using that experimental setup.To date, only a handful of markers are considered completely specific, and most antibodies are also associated with diagnostic pitfalls.Thus, pathologists should refrain from using large screening panels of antibodies because that is likely to result in faulty diagnoses.
Cytogenetics.Karyotyping (analysis of the chromosomal composition of cells) has played a significant role in understanding the genetic abnormalities associated with sarcomas.One of the breakthrough discoveries in karyotyping came in the early 1970s, when the translocation t(9;22)(q22;q12) identified in chronic myeloid leukemia was characterized [35].This finding paved the way for the recognition of specific translocations as diagnostic markers for certain sarcomas.For example, the identification of the translocation between chromosomes 11 and 22 (t(11;22)(q24;q12)) in Ewing sarcoma in 1983 was a significant milestone in understanding the genetic basis of this particular bone sarcoma subtype [36].
Comparative genomic hybridization, in which gene doses of tumor DNA are compared to those from a reference DNA, generates a comprehensive picture of gains and losses (copy number) of genetic material throughout the genome.With a higher resolution than karyotyping, it can give additional information about amplifications of specific genes.However, small gains and losses of genetic material, as well as balanced chromosomal translocations, or single nucleotide variants, will not be picked up by genomic hybridization [37].Fluorescence in situ hybridization (FISH) is widely used in routine diagnostics and can determine specific chromosomal rearrangements or copy number changes of a target region [38].
Molecular analyses.Polymerase chain reaction is a methodological cornerstone of molecular biology.
In clinical practice, it can be used to amplify and sequence genes or identify tumor-specific alterations such as fusion genes.While inexpensive and quick, the method is limited to studying small regions or specific events.For fusion genes, the specific genomic break point must be known.The exact base pairs which merge in the fusion gene may be different in different tumors [39].In recent years, with the emergence and development of new sequencing technologies, comprehensive genomic profiling has become accessible in routine health care in middle-and high-income countries.This opens a new chapter in the mapping and understanding of the sarcoma genome.
Whole-genome, whole-exome, and whole transcriptome sequencing (WGS, WES, and WTS) are high-throughput techniques to read the whole genome, the coding parts (exons), or the transcribed genome (RNA), respectively.Billions of sequencing reactions are carried out in parallel, which is also referred to as massive-parallel sequencing or next-generation sequencing.The main steps for WGS, WES, and WTS are DNA or RNA isolation, library preparation (creation of the genetic base pair sequence representing the input genome), sequencing and imaging (registration of sequence of the bases), bioinformatics (processing of the sequence reads), variant filtration, and variant annotation (detection of potential variants of interest) [40].During library preparation, genomic regions of interest can be selected (targeted or panel sequencing).A growing number of massive-parallel sequencing panels have been introduced in clinical practice [41], but the content targeted is often limited to variants found in carcinomas.Though still costly, WGS has the potential to replace karyotyping, comparative genomic hybridization, FISH, and Sanger sequencing in a single analysis (Fig. 1).However, mapping epigenetic changes-such as methylation statuscurrently requires additional techniques [42].

Massive-parallel sequencing in cancer
In cancer care, genetic analyses are widely used both for diagnostics, targeted therapy decisions, and prognostic predictions.Driver events (genetic aberrations leading to increased malignant potential, driving the cell into a cancer cell) are known for many cancer types.The list of applications for genetic analyses in cancer is long.It ranges from single nucleotide variants in tyrosine kinase receptors (determining sensitivity to tyrosine kinase inhibitors), to more complex analyses of mutational patterns (for instance homologous repair deficient tumors which may respond to PARP inhibitor treatment) [43].Over the last years, the number of clinically actionable genomic biomarkers has rapidly increased both due to an increased knowledge of the cancer genome, but also the demand for specific biomarkers for new and expensive cancer drugs.In a study of 1200 patients, around 70% of those with a successful WGS analysis were found to have one or more actionable biomarkers, with potential therapy options [44].However, the genetic etiology of sarcomas as a group is different from carcinomas and they are perhaps not likely to harbor clinically actionable variants as frequently [13,45,46].The clinical utility of broad genomic profiling in sarcomas is therefore likely to have a bigger impact on tumor classification.In a study by Gounder et al., 10.5% (N = 789) of sarcomas were reclassified after genetic profiling with targeted massiveparallel sequencing [47].Similarly, WGS has been shown to revise 3%-14% of sarcoma diagnoses [13,42,48].In 1310 rare cancers-soft-tissue sarcomas being the largest subgroup-Horak et al. found that molecular profiling, including WGS in some cases, followed by molecular tumor board meeting led to diagnostic reevaluation, genetic counseling, and/or recommendations for experimental therapy in 88% of the cases.In the cases receiving targeted therapy based on the molecular profiling, the overall survival was significantly improved.The benefit from this approach was not clear for all tumor subtypes, though.For instance, the molecular tumor board recommendations did not improve the detection of actionable biomarkers in gastrointestinal stromal tumors (GISTs) [49].Vingiani et al. reported that a molecular tumor board approach led to 9% of the included cohort receiving personalized therapy [50].There is a large discrepancy at the country level on WGS tumor analysis reimbursement.Some countries lack access to it [51], whereas other countriessuch as the United Kingdom-provide wide access to genomic testing with broad genetic analyses of sarcomas and some other cancers [52].

Molecular diagnostics
Verification of specific genomic alterations using molecular methods or immunohistochemistry is compulsory or recommended for some sarcoma diagnoses [18].These alterations are strongly associated with specific sarcoma diagnoses, and their detection supports the diagnosis of a subtype.Examples of molecular markers which strongly support a diagnosis are shown in Table 1.These are often not routinely analyzed outside research trials, but guided tests can be of benefit in cases of unclear diagnosis.A list of molecular genetics, cytogenetics, and associated hereditary syndromes for sarcoma diagnoses with molecular markers that have impact on diagnostics-as described in the WHO Blue Book for soft tissue and bone tumors [18]-is available in Table SA.Most genomic variants are not completely pathognomonic, and genetic findings must always be put into the clinical and histological diagnostic context.Notably, the EWSR1 gene is known to be quite promiscuous and may be rearranged with multiple partner genes.The fusion gene can have significant diagnostic impact, with certain pitfalls.For example, the EWSR1::ATF1 fusion has been described in clear cell sarcomas [53], primary pulmonary myxoid sarcoma [54], and angiomatoid fibrous histiocytoma [55], but also mesothelioma [56] and salivary clear cell carcinomas [55].Figure 2 depicts the COL1A1::PDGFB gene fusion schematically, an example of a gene fusion strongly supportive of the entities dermatofibrosarcoma protuberans, but also giant cell fibroblastoma (Table 1).[58], or distinguishing desmoplastic fibroma of bone from fibrous dysplasia and low-grade central osteosarcoma, by excluding GNAS pathogenic variants and MDM2 amplification [18].
Most sarcomas have low mutational burden (e.g., few single nucleotide variants) as compared to carcinomas and rarely exhibit microsatellite instability [63].Complex karyotypes are present to various degrees in undifferentiated pleomorphic sarcoma, malignant peripheral nerve sheath tumor, and leiomyosarcoma.They often harbor biallelic variants or homozygous deletions in tumor suppressor genes, such as TP53, RB1, ATRX, and PTEN [64].Steele et al. described extreme copy number heterogeneity, pseudohaploidization, and polyploidization as key characteristics of undifferentiated sarcomas, often with co-occurring biallelic loss of crucial tumor suppressor genes or amplification of oncogenes [65].Notably, the p53 and Rb1 pathways are also implied in several of the oncogene-driven sarcomas, including subtypes with aberrant activation of MDM2, CDK4, TERT, CCND1, MYC, and MET [42,66,67].Conventional osteosarcomas often show inactivation of tumor suppressor genes-including TP53, RB1, CDKN2A, and PTEN-and are characterized by genetic heterogeneity and chromosomal instability, aneuploidy, and chromothripsis (a catastrophic event in which a chromosome is broken into thousands of pieces and put back together rearranged) [68].
Recent studies have demonstrated a wide range of genetic driver aberrations in osteosarcoma, with over 30% of osteosarcoma patients harboring potentially actionable targeted variants, most commonly CDKN2A deletions [69].Other subtypes of osteosarcoma-including low-grade central osteosarcomas and periosteal osteosarcomasharbor amplification of the MDM2 and CDK4 genes, which may be helpful in differential diagnostics.
Combined with low karyotypic complexity, the amplification is characteristic of low-grade central osteosarcoma or well-differentiated liposarcoma [18], whereas the presence of a complex karyotype supports dedifferentiation.This amplification is also characteristic of liposarcomas and can help differentiate against benign lipomatous tumors [32,70].
The interpretation of genome-wide copy number variant (CNV) patterns has been proposed to have diagnostic utility but is still not widely used, probably due to the complexity in interpreting the results [71].Specific CNVs or complexity characteristics can help differentiate between diagnoses; for example, the CNV profile differs pleomorphic rhabdomyosarcoma from alveolar or embryonal rhabdomyosarcoma, a distinction that is essential when choosing the appropriate treatment plan [72].An increased understanding of the mechanisms underlying the genomic complexity and heterogeneity is important for the development of novel targets and improving the prognosis of both bone and soft-tissue sarcomas.
Although many sarcoma subtypes are associated with specific fusion genes or events, the majority are poorly differentiated tumors with pleomorphic cell morphology.They are characterized by complex, large structural alterations on the chromosome level, accompanied by inactivation of genes associated with genomic stability.Importantly, complex structural alterations frequently result in the generation of fusion gene transcripts that can be detected using RNA sequencing, but these events are frequently stochastic in nature and not important tumor drivers [73].Thus, identification of novel fusion genes must be interpreted with caution.

Predictive biomarkers and therapeutic targets
Limited knowledge about potential molecular targets and the very limited success of studies investigating targeted therapies are partly related to the curbed therapeutic advances for sarcomas in the past 40 years.A "one-size-fits-all" strategyoften due to the rarity of the condition-is still largely applied, despite known biological differences in different subtypes [74].Indeed, for the majority of soft-tissue sarcomas, anthracyclines remain the treatment of choice in the early setting and as first-line chemotherapy in metastatic disease [11].Bone sarcomas are treated with a combination of chemotherapeutic drugs depending on their histological subtype-with osteogenic sarcomas receiving methotrexate, cisplatin, and doxorubicin, and Ewing sarcomas receiving a combination therapy of five drugs (vincristine, cyclophosphamide, doxorubicin, ifosfamide, and etoposide) [75,76].Although the frequency and composition of the different chemotherapeutics have been changing in Ewing sarcoma, the same agents have been used for the last decades in bone sarcomas [77].
Identification of biomarkers predicting treatment response is an urgent need in sarcomas.In particular, in soft-tissue sarcomas in which the response to chemotherapy is variable, an improved selection of patients would be of great value.Today, certain molecular biomarkers are in clinical use (Table 2).The majority of these predict responses to targeted therapy, whereas established biomarkers for chemotherapy response in general are lacking.Biomarker selection of treatment improves the clinical outcome in comparison to an unselected treatment approach [78].Targeted therapies are clinically available and used worldwide for some sarcoma diagnoses, such as GIST, and locally aggressive benign mesenchymal tumors such as tenosynovial giant cell tumors and inflammatory myofibroblastic tumors [79,80].The gain-offunction KIT or PDGFRA variants in GISTs serve as predictive indicators of tyrosine kinase inhibitor treatments, except the imatinib-resistant PDGFRA D842V variant [81][82][83].For the latter, prognosis has been improved since precision medicine has led to identification of the novel tyrosine kinase inhibitor, avapritinib [84].Moreover, patients with benign tenosynovial giant cell tumors with elevated levels of CSF1-caused by translocation/fusion and other not fully known mechanisms-can often be offered pexidartinib or vimseltinib [85][86][87], an example of an emerging molecular marker for sarcoma treatment.
Unfortunately, molecular driver events-enabling targeted therapies-are not known for most sarcoma diagnoses.The role for molecular diagnostics as a guide for treatment has thereby been limited until now [88], and therapeutic targets in sarcomas except GIST are mostly the ones that are applicable to all solid tumors (Table SB).For example, NTRK gene fusions can be found in several sarcoma subtypes, and result in oncogenic TRK activation [89].Sarcomas that harbor NTRK fusions show excellent response to TRK inhibitors [90,91].To this extent, expert recommendations have been developed to guide screening and management [92].Some sarcoma-specific biomarkers and therapeutic targets are emerging with compelling clinical and biological evidence, such as SMARCB1 and MDM2/CDK4 (Table 2).
Even though the number of targeted therapies for sarcoma subtypes is relatively low, there are molecular profiles that could benefit from targeted treatments regardless of histopathological diagnosis.For example, tumors with high tumor mutational burden (somatic genetic variants per coding cancer genome) could benefit from immune checkpoint inhibitor treatment.Tumor mutational burden is not a major cancer driver event in sarcomas, and therefore not part of the standard diagnostic workup.It has been seen primarily in microsatellite instable tumors [63].However, tumor mutational burden can occasionally be high in some sarcomas, such as in leiomyosarcoma, myxofibrosarcoma, undifferentiated pleomorphic sarcoma, alveolar soft part sarcoma, and cutaneous angiosarcoma [134][135][136][137]. High tumor mutational burden could change the recommended treatment strategy, and, supposedly, testing for it could be prioritized in these subgroups, when prioritizing is necessary [138].

Prognostic biomarkers
Numerous studies have identified prognostic molecular markers in different sarcoma subtypes.
Most of these markers were investigated in relatively small cohorts and have never been validated.
For even fewer, their independent prognostic significance when considering established prognostic factors has been demonstrated.As an example, in solitary fibrous tumors, there are established and validated recurrence risk classification systems [139].It has been shown that the type of NAB2::STAT6 fusion in solitary fibrous tumors is associated with outcome.Fusions containing only the transactivating domain of STAT6 display a shorter recurrence-free and overall survival [140], but whether the fusion variant predicts outcome within each risk group and adds additional prognostic information is unknown.This illustrates that clinical utility, ideally demonstrated in prospective trials, is a prerequisite for the development of prognostic molecular biomarkers for clinical use.
In GIST, the type of KIT or PDGFRA variant carries both predictive and prognostic information.PDGFRA-mutated tumors generally have a better prognosis, whereas tumors with KIT variants involving codons 557 and 558 show an aggressive behavior [141,142].In rhabdomyosarcoma, tumors with PAX3::FOXO1 fusion have a worse outcome compared to those with PAX7::FOXO1 fusion [143].In desmoid fibromatosis, the type of CTNNB1 pathogenic variant was thought to be associated with prognosis, with the S45F variant being a high-risk factor for recurrence [144].However, it has rather been shown that the CTNNB1 genetic profile is associated with unfavorable prognostic factors, and not a prognostic factor in itself [145,146].
Genomic complexity is associated with poor outcome across several sarcoma subtypes [42,47,147].Genomic complexity usually refers to structural alterations that arise through various biological mechanisms, such as whole-genome doubling, aneuploidy, loss of heterozygosity, and chromothripsis.There is no consensus on how to measure the level of genomic complexity in sarcoma, and it is unknown which pattern of copy number alteration is most closely linked to the clinical behavior within each histological subtype.The gene expression signature CINSARC (complexity index in sarcomas) is probably the most studied marker of genomic complexity [147], and its utility is currently being investigated prospectively [148] in different trials (NCT02789384, NCT04307277, NCT03805022).

Discussion
The molecular profiles for sarcomas differ substantially, ranging from highly complex karyotypes to only one fusion gene event driving the malignant process [18].However, for most sarcoma subgroups, the molecular profiles and cancer-driving mechanisms are unknown.There is a need for deeper understanding of the sarcoma genome.In this review, we describe the diagnostic, therapeutic, and prognostic molecular markers that are used in the clinic or are expected to enter in the near future.
Given the risk for undesirable adverse events, personalizing therapy-and selecting patients with higher probability of response and preventing those who probably will not benefit from it-is of major importance.Moreover, prognostic and predictive biomarkers can improve patient selection in clinical trials, increasing the probability of identifying successful interventions and making it less likely to dilute potential effect.Last but not least, on a health-economic level, the increased cost of oncology drugs demands companion diagnostics, where molecular or other tests determine the likelihood of therapy efficacy.Unfortunately, the economical reimbursement for diagnostic markers is lagging behind, effectively limiting the implementation of new and comparably expensive methods.That said, it is important to evaluate and determine the clinical benefit and cost reduction of a precise diagnosis for soft tissue and bone tumors.
Given their rarity and the potential detrimental consequences for the individual in the event of faulty diagnosis, sarcoma diagnostics would be a natural starting point for the implementation of comprehensive diagnostic genomics.
To evaluate the implementation of new techniques in health care, a health technology assessment (HTA) is preferably used.This approach requires real-world data and could be a basis for strategic consensus decisions [149].Moreover, additional decision models-such as value of information Fig. 3 Potential implications from unsolicited genetic findings after massive-parallel sequencing.Genetic screening can generate unsolicited germline findings (hereditary variants associated to syndromes with or without cancer predisposition).These variants need to be interpreted in the clinical context.If the variants are considered clinically relevant, carrier testing can be offered in the family.Carriers may, for some syndromes, benefit from surveillance programs, riskreducing surgery, and prenatal testing counseling.Moreover, this may have an impact on the treatment options and clinical follow-up for the patient.However, if the unsolicited findings are variants without a known risk profile, there might not be any robust recommendations for clinical follow-up.This can be complicated in the genetic counseling situation, which lead to unnecessary anxiety, and the question of whether to offer genetical tests to other family members or not (segregation analysis; a way of increasing or decreasing the probability that a genetic variant is indeed associated to a syndrome or trait).
analysis with integrated scenario drafting-can be used when such data is missing.Simons et al. performed an investigation of the costeffectiveness, budget impact, and impact of the uncertainty of future developments for clinical use of WGS compared with standard of care in patients with lung cancer [150].They conclude that WGS may likely become cost-effective, provided it identifies more patients with actionable targets.Large international programs implementing WGS broadly in oncology diagnostics are ongoing and will add real-world data to either support or dispute its economic benefit.In the Netherlands, implementation of WGS diagnostics for all metastasized solid tumors is performed, including the secondary endpoint of HTA, with micro-costing and budget impact analysis [151].No HTA or evaluation of economic benefits from WGS diagnostics in sarcomas has yet been published.However, ongoing projects such as the molecular profiling of advanced soft-tissue sarcomas study and the 100, 000 Genomes Project aim to provide evidence supporting the implementation of WGS in clinical practice for soft-tissue sarcomas [13,14].These randomized multicenter trials might enable future HTAs for sarcomas.
Unsolicited findings (Fig. 3) should be expected from genetic testing, and according to the European Society of Human Genetics, "only genes with a known (i.e., published and confirmed) relationship between the aberrant genotype and the pathology should be included in the analysis" [152].This statement applies mainly to diagnostics of hereditary disorders, but in general, the reasoning is applicable also to clinical somatic genetic analyses.Somatic analyses might generate knowledge about the germline, and a hereditary disease that has no association with the cancer might be revealed.This is especially true when a healthy tissue sample (such as blood) is not analyzed in parallel with the tumor.In addition, it has been observed that, in sarcoma patient cohorts, there might be an enrichment for cancer syndromes not previously known to be associated with sarcomas [153][154][155].There is usually no consensus on surveillance recommendations or prophylaxis for these syndromes with regard to the sarcoma development risk.These findings have implications for both the patients and their relatives.Patient information about the possible outcomes from cancer genome sequencing is of importance, and there should be a careful action plan for when unsolicited findings are encountered.

Future perspectives
Analysis of liquid biopsies-for instance, cell-free DNA-is an emerging field in cancer care.Cell-free DNA is short fragments of DNA released into a body fluid, such as blood, which represents the genome of the cell releasing it.Cell-free DNA can be isolated from a blood sample, and when a malignant tumor is present in the body, some of the cell-free DNA comes from the cancer cells [156].
There is a growing number of studies describing how analysis of cell-free DNA can be utilized for sarcoma diagnostics [157], treatment monitoring [158], identification of resistance mechanisms [159], and minimal residual disease detection [160].This application is bound to grow rapidly.
The insight into sarcoma epigenetics is also improving [42].An increasing number of mesenchymal tumors are considered to be mechanistically determined by epigenetic changes in the genome.They have widespread epigenetic dysregulation, which in many cases is caused by a small number of genetic changes.For instance, chondrosarcomas are driven by methylation dysregulation [161]; giant cell tumors of bone have histone modifications [162]; Ewing sarcoma fusions affect higher order chromatin organization [163][164][165]; and the fusion gene in synovial sarcomas alter the BAF-complex, resulting in a large epigenetic reprograming of the cell [33].Global DNA methylation patterns have become central in the classification of central nervous tumors [166].
Similarly, methylation patterns can be used in sarcoma classification, but the utility has not yet reached the same level.
Molecular diagnostics has become an important instrument in bone-and soft-tissue tumor classification and a significant tool to determine treatment prognostics for a few subtypes.With the exploding availability of sequencing technologies, it becomes increasingly important to understand the strengths and weaknesses of these methods.Perhaps most importantly, a well-trained pathologist is required to interpret genomic findings in the context of tumor histomorphology.We foresee that a majority of sarcoma treatment centers will increase the use of massive-parallel sequencing, and that the requirement of genetic validation is likely to increase.Hopefully, the growing number of treatment prognostic biomarkers in oncology will continue to grow, allowing patients with sarcoma to benefit from targeted therapies.

Fig. 1
Fig. 1 Overview of common genetic analyses in sarcoma diagnostics.Presentation of some of the most common genetic analyses performed in the clinical setting.A karyotype has the lowest resolution, detecting genetic aberrations of approximately 10 mega base pairs, while sequencing (such as Sanger sequencing, and wholegenome or whole-exome sequencing) can detect aberrations of one single base pair.Depending on the diagnostic genetic change sought, different methods can be used.Genomic arrays: for instance, comparative genomic hybridization.FISH, fluorescence in situ hybridization; PCR, polymerase chain reaction.

Table 1 .
Diagnostic molecular biomarkers for soft tissue and bone tumors.

Table 1 .
(Continued) [18]: Data based on the WHO Blue Book for soft tissue and bone tumors[18].Molecular events included in this table are strongly supportive of a diagnosis.For all molecular events mentioned in association with a specific sarcoma diagnosis in the WHO Blue Book, which can be considered supportive of a diagnosis, please see TableSA.Abbreviations: NTRK, neurotrophic receptor tyrosine kinase; SNV: single nucleotide variant; WHO: World Health Organization.Fig.

2 Schematic picture of the COL1A1::PDGFB fusion gene.
This COL1A1::PDGRB fusion gene has been detected in a case with dermatofibrosarcoma protuberans (ongoing study by authors).It is the product of break points within the COL1A1 gene (on chromosome 17) and the PDGFB gene (on chromosome 22), and the fusion of these break points.The detection of this fusion gene is strongly supportive of the diagnosis.

Table 2 .
Sarcoma-specific molecular biomarkers, FDA level 2-3.Treatment prognostic biomarkers for soft tissue and bone tumors.Data extracted from the OncoKB database on 2023-07-01.The table includes all FDA approved biomarkers.See TablesSB1 and SB2for agnostic molecular biomarkers for sarcomas.
Abbreviations: FDA, Federal Drug Association; GIST, gastrointestinal stromal tumor; MSI, microsatellite instability; TMB, tumor mutational burden.a There are numerous variants with varying prognostic value to predict sensitivity and resistance to TKIs.