Ewing's sarcoma (ES) is the secondly common malignant bone and soft tissue tumor in children and adolescents, with an incidence of 1 in 1.5 million. It is a small round cell malignant tumor originating from bone marrow mesenchymal stem cells. Eighty-five to ninety percent of ES patients have typical chromosomal translocations [t_{(11;22)(q24;q12)}] and the expression of EWS-FLI1. It is characterized by young onset age, non-specific clinical features, strong invasion ability and poor clinical prognosis (1–5). ES often occurs in the epiphysis of the extremities, mainly presenting as swelling with palpable mass and mild pain, accompanied by obvious night pain and pathological fracture, which seriously affects the quality of life of patients. Early diagnosis lacks specificity with a high misdiagnosis rate (5–8).

Patients with ES have poor long-term prognosis in spite of the variety of treatments and the progress made in surgery, radiotherapy and chemotherapy (1, 9, 10). Up to 80% of patients with ES have distant metastasis at initial diagnosis, whose overall 5-year survival rate is <30%. Although the application of surgery and multi-agent neoadjuvant chemotherapy has resulted in a significant improvement in the prognosis of patients with ES, the survival prognosis of patients with recurrence and metastasis remains poor (11, 12).

Few previous studies have involved combining multiple independent factors to predict lymphatic metastasis in patients with ES (13, 14). Therefore, accurate prediction of the probability of lymphatic metastasis has high application value. As a prediction model based on independent factors of clinical prognosis, nomogram has good stability and applicability (15, 16). Therefore, it has been widely used to evaluate tumor prognosis and metastasis (17–21). Likewise, it can be used to predict and evaluate the prognosis of patients with ES. Besides, it can help improve early detection, disease surveillance, overall patient survival rate and their life quality.

In this study, the patients' data was extracted from the SEER database in order to establish a nomogram. Furthermore, we validated it externally via multi-center clinical data of patients. To provide patients with more accurate medical services, we also designed a web calculator that could help clinicians quickly and accurately assess the risk of lymphatic metastasis in various ES patients.

Ewing's sarcoma (ES), first reported by James Ewing in 1921, is an aggressive bone and soft-tissue tumor (5, 9, 22, 23). Thanks to a variety of treatment measures have been adopted in ES, the prognosis of patients with localized forms of the tumor has been significantly improved. However, the 5-year survival rate for metastasis in patients is still lower than 30% (5). Therefore, it is essential to seek more effective treatment methods to prolong the survival time of patients and improve the prognosis of patients. Some scholars pointed out that the clinical management and prognosis of ES could benefit from the biological markers, nevertheless, there was no biomarker that could really predict the prognosis of patients (24). Advancements in artificial intelligence (AI) and machine learning approach can also help clinicians with the timely diagnosis and more efficient treatment of patients in the healthcare sector (25–27). Nomogram was much more convenient to obtain the predictors. Hence, it was considered a big advantage over traditional staging, which had been proposed as an alternative and even a new standard to guide the treatment of cancer patients (17, 28, 29). We screened the clinical information of 929 patients with ES using a SEER database and found that survival time, race, T stage, M stage, surgery, and lung metastasis were significant factors that affected lymphatic metastasis in ES patients. In view of the prognostic factors above, the integration of the nomogram could be intuitively used in the clinical evaluation of lymphatic metastasis in patients with ES. After the model was established, we collected the data of 51 patients from four clinical centers and conducted external validation to confirm the reliability of the model, thus enhancing the value of the model in clinical application. Besides, we established a web calculator for the prediction and evaluation of lymphatic metastasis in ES patients.

Logistic regression analysis indicated that T stage, M stage and survival time were independent factors for lymphatic metastasis in ES patients. Several consistent prognostic factors were identified in North America and Europe. Multiple studies showed that in addition to stage, poor prognostic factors included tumor size and metastasis at diagnosis (30–32), which were basically consistent with our findings.

Tumor size was considered to be an important prognostic factor. Large initial tumor size has been repeatedly identified as negatively affecting survival in ES patients (7). The correlation between tumor size and metastasis remains to be further studied. We hypothesized that the larger tumor size may have more potential invasion and metastasis, and the existence of abundant lymphatic tissue also promotes the occurrence of lymphatic metastasis.

The current study also showed that most patients began with micro-metastases (33). Furthermore, ES had a wide range of metastasis sites, including lung, bone, lymph nodes, liver and brain (1). Bone and lung metastases and the development of extensive tumors were linked with poor prognosis. Patients were most likely to develop lung metastases than bone metastases, and patients with isolated lung metastases had a better prognosis than those with extra-pulmonary metastases. Patients with unilateral lung involvement appeared better than those with bilateral lung (34). Our study also revealed that M stage was a key predictor. In this study, about 59% of patients with lymphatic metastasis developed the state of M1, and the OR of lymphatic metastasis in M1 patients was 2.061. Moreover, patients with lymphatic metastasis have a higher tendency to develop lung metastasis, which is consistent with the results of logistics regression analysis. For the risk of metastasis of the confirmed patients, systemic staging can be performed with CT and MRI for primary site, and ¹⁸F-fluorodeoxyglucose(FDG)PET-CT scan.

Based on previous studies, ES was more commonly found in men than in women (sex ratio of 3:2), and the prevalence was over seven times higher in whites than in blacks (1, 5, 35). In addition, Duchman et al. found that blacks had an increased frequency of metastases when diagnosed with ES compared to patients of other races (36). Our results showed that the incidence of lymphatic metastasis was more than 2 times higher in black patients than in white patients, and 1.3 times higher in men than in women.

Although this model provided theoretical support for clinical diagnosis and treatment, some shortcomings remained in our study: (1) first of all, although we established the lymphatic metastasis in patients with clinical prediction model which was verified by external cohort, external validation group of small number of cases might be a problem such as selective bias; (2) due to incomplete data entry in SEER database, we were unable to analyze all factors that might be associated with lymphatic metastasis. For instance, it was lack of some laboratory indicators, such as serum lactate dehydrogenase (LDH) levels. Based on this, we could avoid bias by further increasing the number of cases in the external cohort. In addition, we could use the case data of Asian people to establish a more suitable prediction model for Asian people so that we can make a comparison of this model.

Our study established the associated nomogram by using data from the SEER database and comprehensively evaluated predictors of lymphatic metastasis in ES patients via external validation. The prediction of the risk of lymphatic metastasis could guide clinicians to carry out individualized and precise treatment for patients, and help to formulate follow-up treatment measures and plans for patients.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

All participants signed written informed consents following the recommendations of the Institutional Review Board of the Xianyang Central Hospital. The ethical approval certificate is 20210022.

CY and ZL designed the study. WeL and ZH performed the study and analyzed the data. WeL and CX wrote the manuscript. SD, HW, QL, and Z-RT provided the expert consultations and clinical suggestions. WaL participated in its design and coordination. BW helped to draft the manuscript. All authors reviewed the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.