Rhabdomyosarcoma (RMS) is a rare but aggressive malignant tumor originating from the embryonic mesenchyme of the rhabdomyosarcoma tissue (1) and is the most common soft tissue sarcoma (STS) in children and adolescents (2), accounting for about 4%–5% of cancers in the pediatric population (3). Although RMS can occur anywhere in the body, 40% of primary lesions originate in the head and neck region that can be further classified as the orbital RMS (OR), parameningeal RMS (PR) and non-orbital-parameningeal RMS (NOPR) (4–6). RMS is characterized by unclear prognostic factors and poor survival, it is reported that high-risk RMS and recurrent disease have 5-year survival rates of lessen 30% and 17%, respectively (7). However, the overall survival (OS) rate of patients has been increased through better staging, risk grouping, improved localized treatments, and better supportive care after The Intergroup Rhabdomyosarcoma Study (IRS) was established in the United States in 1972 (8, 9). The treatment method of RMS is multimodal, involves multi-agent systemic chemotherapy, along with surgical resection of the tumor with or without addition of the radiation therapy for control of local lesion (10). The identification of prognostic factors for RMS may help with optimizing treatment protocols.

A nomogram is a graphical tool for displaying the predicted value of individual survival based on significant variables produced by multivariate regression analysis briefly and intuitively (11, 12); currently, it has been widely used in disease prediction of malignant tumors such as laryngeal carcinoma and lung cancer etc (13–15). There are also some nomograms that have been constructed and proven to be beneficial in the management of RMS (16–18), and they present obvious advantages such as improved predictive accuracy, robustness, and usability, all of which increase their potentials in clinical practices. As a result, nomograms have been proposed as alternative methods, or perhaps new standard, to guide the management of cancer patients (19, 20). However, there is still a lack of nomogram model to predict the survival probability of head and neck rhabdomyosarcoma (HNRMS) patients in children and adolescents, which may be attributed to limited cases reports and small case studies in each single institution.

In this study, we analyzed the demographic, clinical characteristics and treatment strategies for predicting OS of HNRMS patients in children and adolescence by using population-based data recorded in the Surveillance, Epidemiology, and End Results (SEER), with the purpose of providing clinicians and patients with a simple but effective clinical tool to predict survival.

The tumor primary site was classified into favorable and unfavorable based on criteria used for staging pediatric tumors (21). In terms of HNRMS, favorable site refers to OR and NOPR, while unfavorable refers to PR. In our study, both univariate (HR: 7.85, 95% CI: 3.24–19.03, P < 0.001) and multivariate Cox regression (HR: 3.32, 95% CI: 1.09–10.13, P = 0.035) showed that PR was associated with poor OS, which is consistent with previous studies (22, 23). This may be due to the fact that the PR is close to the intracranial area and could not be completely removed (24). Furthermore, compared with superficial sites such as OR, PR is less likely to be detected, making it impossible to find and diagnose the tumor in time, and the lesion may have distant metastasis in early stage, which may metastasize first to the lungs or the bones (25). Orbital RMS affects eye movements at an early stage, so it is usually detected earlier and has a better prognosis.

In our study, the proportion of tumor size more than 5 cm was about 29.4%, which was lower than 51% in the IRS-IV study (9), 47.2% in the study of Ma et al. (26), and 75.5% reported by the Japanese researchers Hosoi et al. (27); this may be attributed to HNRMS limited growth space. Unsal et al. (28) and Dantonello et al. (29) reported that tumor size is a risk factor for poor survival. This is consistent with our study finding that a larger tumor was an adverse prognostic factor in HNRMS patients. Studied showed that tumor size with more than 5 cm was independent risk predictor for survival and local treatment failure (30); the result of univariate Cox regression showed that tumor size with more than 5cm was a risk factor for prognosis in children and adolescence (HR: 3.99, 95% CI: 2.36–6.72, P < 0.001).

RMS was divided into four subgroups: embryonal, alveolar, pleomorphic, and spindle cell/sclerosing RMS, according to the World Health Organization (WHO) classification system, of which embryonal was the most common subtype (31), embryonal RMS had a clearly more favorable clinical prognosis than alveolar type, which was observed to occur at an earlier age and associated with better prognosis (32, 33). Alveolar RMS was common in the muscle of the trunk and limb extremities and presented as a poor prognosis (32, 34, 35), studies showed that microarray-based gene expression profiling showed that PAX-FOXO1 fusion gene-positive alveolar RMS had a worse biological behavior than PAX-FOXO1 fusion gene-negative (36, 37). With this, it was evident that FOXO1 fusion status was an important prognostic factor and has replaced pathologic staging in RMS risk stratification in recent years (38, 39). In our study, univariate (HR: 2.67, 95% CI: 1.67–4.28, P < 0.001) and multivariate Cox regression (HR: 1.82, 95% CI: 1.01–3.02, P = 0.020) showed that alveolar RMS was associated with poorer OS than embryonal RMS, which is consistent with previous research (40).

Distant metastasis of RMS was found to be associated with worse survival; a study conducted by Turner and Richmon (41) found that patients with localized and regional disease had a higher 5-year survival rate than patients with distant metastasis (92.4%, 60.1%, vs. 36.6%), and the HR values were 0.15 and 0.37, respectively. Our results are consistent with Turner and his partner’ findings with the data of univariate Cox regression (HR: 8.23, 95% CI: 4.88–13.87, P < 0.001) and multivariate Cox regression (HR: 2.20, 95% CI: 1.03-4.69, P = 0.041). In addition, exiting literature suggested that tumors of parameningeal sites present advanced stage; they tended to invade distant tissues and organs (34, 42). As mentioned, embryonal RMS occurred with localized disease and in children at an earlier age. On the other hand, alveolar RMS presented most commonly with distant metastasis as well as in adolescents.

Surgical treatment was often consider as a critical treatment means for improving survival of RMS patients (43); surgeons tended to perform radical resection to remove tumor as well as its adjacent tissues and lymph nodes. However, expansion of the surgical scope in pursuit of negative margins would destroy important structures and functions for HNRMS in children and adolescents. Furthermore, tumor sites in dangerous site, such as intracranial extension, bony erosion of cranial base, or cranial nerve palsy, had increased tremendously surgical feasibility (9). The use of radiation and chemotherapy therapy was also found to be effective in improving survival (40, 41); studies showed that radiotherapy combined with chemotherapy is primary means for patients with unresectable primary tumors after initial treatment (44, 45). Our study showed that surgery, radiotherapy, and chemotherapy were all effective treatment methods for HNRMS in children and adolescences (P < 0.001, P = 0.013, P < 0.001, respectively).

There are a few limitations of this study. First, there are several editions of TNM stage, which could have an affection on IRS stage. Second, the SEER database lacks information on laboratory tests, specific radiotherapy and chemotherapy regimens, and comorbidities; however, they also play a critical role in prognosis of HNRMS (46, 47). Szkandera et al. (48) demonstrated a decreased lymphocyte/monocyte ratio (LMR) represents a novel independent poor prognostic factor in STS patients. Fausti et al. (49) also drew consistent conclusions that LMR is a specific predictor of Trabectedin efficacy and could be useful in daily clinical practice. Moreover, they highlighted a possible correlation between LMR levels and the percentage of intratumoral macrophages. Finally, this is a retrospective study, and the survival outcomes reported in SEER database are associated with different treatment guidelines, which are different from every era. Thus, selection bias could not be avoided.

A prognostic nomogram model was constructed for predicting the probability of survival in HNRMS in children and adolescents, and this model presented practical and applicable clinical value to predict survival when choosing treatment strategies.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

JW: Data curation, Software, Visualization, Writing – original draft, Writing – review & editing. QZ: Funding acquisition, Resources, Supervision, Writing – original draft, Writing – review & editing.