Second Malignant Neoplasms in Patients with Rhabdomyosarcoma: A US Population-Based Analysis


 Objective

Rhabdomyosarcoma (RMS) is a rare malignant tumor. The main treatment modality is comprehensive with chemotherapy, radiation therapy, and surgery. With the advancement in recent decades, patient survival has been prolonged, and long-term complications are attracting increasing attention among both physicians and patients. This study aimed to present the survival of patients with RMS and analyze the risk factors for the development of a second malignant neoplasm (SMN).
Methods

The Surveillance, Epidemiology, and End Results (SEER) Program 18 registry database from 1973 to 2015 of the National Cancer Institute of the United States was used for the survival analyses, and the SEER 9 for the SMN analysis.
Results

The 5-, 10-, and 20-year overall survival rates of the patients with RMS were 45%, 43%, and 33%, respectively. The risk of SMN was significantly higher in patients with RMS compared to the general population (SIR = 1.95, 95% CI: 1.44–2.57, p < 0.001). The risk of developing SMN was increased in multiple locations, including the bones and joint (SIR = 35.25) and soft tissues including the heart (SIR = 22.5), breasts (SIR = 2.10), male genital organs (SIR = 118.14), urinary system (SIR = 2.36), brain (SIR = 9.21), and brain and other nervous system organs (SIR = 8.59). The multivariate analysis indicated that RMS in the limbs and earlier diagnosis time were independent risk factors for the development of SMN. Patients with head and neck (OR = 0.546, 95% CI: 0.313–0.952, p = 0.033) and trunk RMS (OR = 0.322, 95% CI: 0.184–0.564. p < 0.001) and a later diagnosis time were less likely to develop SMN (OR = 0.496, 95% CI: 0.421–0.585, p < 0.001).
Conclusion

This study describes the risk factors associated with the development of SMN in patients with RMS, which is helpful for the personalized screening of high-risk patients with RMS.


Introduction
Rhabdomyosarcoma (RMS) , although rare, is a common childhood cancer and is the most common soft tissue sarcoma among children. The overall incidence among those aged < 20 years is 4.5 cases per million people [1] . In the United States, approximately 350 new cases are recorded each year. According to the data of the Surveillance Epidemiology and End Results (SEER) program, the incidence of RMS varies depending on age and histopathology [2] .
Signi cant advances have been made over the past four decades in the treatment of RMS. The majority of children with the early-stage disease will essentially be cured. Because of the collaborative effort dedicated to clinical trials, the current 3-year overall survival rate of patients with high-risk RMS is approximately 80% and treatment failure is largely attributed to local relapse. Therefore, increasing the local control rate remains challenging in the treatment of RMS [1] .
Increasing the dosage of cytotoxic drugs may lead to prognostic improvement but is accompanied by increased side effects, including long-term ones such as Second malignant neoplasm (SMNs), which have particularly drawn the attention of both physicians and RMS survivors [1] . There is a need to assess the risk of SMN in patients with RMS. However, it is di cult to do so in a single-center cohort with a longterm follow-up due to the limited incidence. Therefore, this study aimed to assess the risk of SMN development in patients with RMS through this population-based study using clinical data from the SEER database.
The survival time used in this study was Overall survival(OS). SMNs were de ned as other malignant tumors that were present at least 2 months after the diagnosis of RMS. Information at diagnosis, including age, sex, race, pathologic type and differentiation level, year of diagnosis, major location, radiation therapy, chemotherapy, survival time, and survival condition was collected. Authorization was obtained from the SEER website to collect the data from the database and no additional ethical approval was required.

Statistical analysis
Categorical data was presented as frequencies and percentages.The Kaplan-Meier method was used for survival analysis. Univariate analyses were performed using the log-rank test, and multivariate using the Cox regression model. Independent risk factors for the development of SMNs were determined using logistic regression analysis. The Standardized incidence ratio(SIR)was calculated using SEER*Stat and was analyzed and presented with 95% Con dence interval CI . All statistical analyses were performed using SPSS 22.0 (SPSS, Inc., Chicago, IL). A p-value of ≤ 0.05 was used to indicate statistical signi cance.

Demographic and clinical characteristics of study patients
A total of 4,787 pathologically diagnosed patients with RMS were included in this study. The demographic and clinical characteristics of all patients with RMS and those with SMNs are shown in Table 1. The trunk was the most common primary site (50.6%, n = 2,420), followed by the head and neck (36.1%, n = 1,728) and the limbs (12.2%, n = 586). In terms of pathologic types, embryonal was the most common (41.0%, n = 1,964), followed by NOS (36.5%, n = 1,746) and alveolar type (22.5%, n = 1,077). The risk of primary SMNs was 34.1% in the head and neck and 32.9% in both the limbs and the pelvis. Patients with SMN were most likely to have NOS (54.7%) and embryonal (36.0%) as opposed to alveolar type.

OS of patients with RMS
The 5-, 10-, and 20-year OS rates for the entire RMS cohort were 45%, 43%, and 33%, respectively. The survival data of the different variables are presented in Fig. 1. Univariate analysis of the independent risk factors for mortality was initially conducted, and those with p < 0.1 were included in the multivariate analysis.

Analysis of SMN risk factors
The univariate analysis showed that the year of diagnosis (p < 0.001), primary site (p < 0.001), and chemotherapy (p = 0.006) were associated with SMN, whereas radiation therapy was not (p = 0.58

Discussion
In this study, we found that the development of SMNs, primary site, pathologic type, age, radiation therapy, and the degree of differentiation were independent risk factors affecting the prognosis of patients with RMS. Compared to the general population, patients with RMS had a signi cantly increased risk of developing SMNs, and risk factors found to be associated with their development included primary site and the time of diagnosis.
In recent decades, survival has signi cantly improved among patients with RMS. This has been mainly due to the multidisciplinary treatment approach, which includes optimal cytotoxic drug combinations, and advances in radiation therapy. The greater number of RMS survivors leads to an increased risk of late complications, such as SMNs. The results of our study suggested that the year of diagnosis was a risk factor affecting the incidence of SMN, and the incidence of SMN was increased with the prolongation of survival time. Heyn et al. found that the incidence rate of SMN increased signi cantly with increased time since diagnosis [3] . Studies conducted on childhood tumors have also reached similar conclusions of the prolonged time of several years, and sometimes even decades of follow-up to observe the occurrence of SMN. These studies showed that the 10-year cumulative incidence rate of SMN in patients with childhood tumors was approximately 1.2-1.7%, the 20-year cumulative incidence rate was 3.3-4.3%, and the 25- year cumulative incidence rate was 3.7-12.1% [3][4][5][6][7][8][9][10] . The above studies indicate that the incidence of SMN will gradually increase as the survival period of childhood tumor survivors increases. We believe that there are two possible reasons. One is that the SMN (such as leukemia, radiation carcinogenesis, etc.) caused by therapeutic factors usually takes years to decades to manifest. The other is that survivors are getting older, and the incidence of SMN is increased due to non-therapeutic factors represented by heredity. The above two factors work together to cause the diagnosis time to become a factor that affects the occurrence of SMN. Moreover, we found that patients with RMS of the limbs were more likely to develop SMNs. There have not been relevant research studies that verify this conclusion, and the effects of different primary sites of RMS on SMN have not been explored. Because RMS of the extremities is correlated with a higher rate of treatment failure, we suppose that the higher risk of SMN of the extremities could be related to the cumulative dose of cytotoxic drugs and the intensity of radiation therapy.
Whether radiotherapy and chemotherapy will increase the incidence of SMN is still controversial. The results of our study did not show that radiation therapy and chemotherapy were independent risk factors for the development of SMNs, similar to other studies that found that the SIR of SMN for patients who received radiation therapy was not signi cantly different to that of patients who did not [5] . In addition, for most types of radiation therapy, the SMN sites were located outside the eld of irradiation. Therefore, it was speculated that radiation therapy did not directly lead to an increase in the incidence of SMN [5] . But the mainstream view is that radiotherapy is a risk factor for SMN [11][12][13][14] . They found that the radiation therapy doses may be correlated with the risk of SMN, Moreover, radiation is carcinogenic and the mechanism is unclear,It is di cult to exclude the in uence of it on SMN. Whether chemotherapy affects independent risk factors for SMN is also controversial, and mainstream views also believe that chemotherapy can cause SMN. Cytotoxic drugs, such as cyclophosphamide and etoposide, have been reported to signi cantly increase the risk of secondary malignant leukemia in patients with RMS [15] . The incidence of SMNs was even lower in patients who used actinomycin for treatment, suggesting that this drug may have some protective effects [15] . Our results showed that chemotherapy was a risk factor for SMN in a univariate analysis (p = 0.006), but it was not statistically signi cant in multivariate analysis (p > 0.05).
Consistent with Scaradavou et al. [5] , our study found that the pathologic type was not a risk factor of SMN. One possible reason is that patients with embryonal RMS are prone to cancer syndromes, such as Li-Fraumeni, which is a correlation that many studies have con rmed [6,[8][9][10] . Li-Fraumeni syndrome is a chromosomal dominant disease that has been closely associated with RMS and with mutations in the tumor suppressor TP53 gene. Common SMNs in patients with RMS include breast cancer, osteosarcoma, brain tumors, leukemia, and adrenocortical malignancy, which are very similar to the clinical characteristics of Li-Fraumeni syndrome. Other studies have also reached similar conclusions [3,5,7] . Since Li-Fraumeni syndrome has a clear family history, the probability that high-risk family members will develop invasive malignant tumors before the age of 30 years is as high as 50%. Therefore, an understanding of the patients' family history is important in order to establish which ones require more frequent and detailed screenings and helps determine the etiology of SMN. Time since diagnosis was found to be another risk factor for the development of SMN.Some studies have suggested that the main factors associated with the development of SMN were internal, such as cancer susceptibility (including a family history of Li-Fraumeni syndrome) rather than external such as radiation therapy and chemotherapy. Another important issue is whether the incidence of SMNs affects both solid tumors and hematological malignancies. It has been believed that the increase in the incidence of SMNs is signi cant in both types, yet the ndings of our study indicate that the SIR for SMNs in all solid tumors was 1.95 (95% CI: 1.54-2.44, p < 0.05), but it was not statistically signi cant in hematological malignancies. In addition, there was an increased risk of some speci c types of hematological malignancies, such as nonlymphocytic leukemia. Heyn et al. showed that the rates of solid SMNs (represented by osteosarcomas) and hematological malignancies (represented by acute non-lymphocytic leukemia) were signi cantly increased in RMS patients [3] , whereas Archer et al. found that the increased risk of SMNs was only signi cant in solid tumors [16] . This could be related to the sample size of the different studies, duration of the follow-up as well as the evolvement of systemic treatment.
This study has several limitations. Firstly, there are biases inherent to the retrospective design. Moreover, detailed treatment information, such as the exposure range of radiation therapy target areas, as well as the chemotherapy regimen and dose could not be acquired from the SEER database. Lastly, information on the tumor size, in addition to the sequence of radiation therapy and chemotherapy was not available.

Conclusion
This study is a population cohort-based study that objectively describes the occurrence of SMN and its risk factors in patients with rhabdomyosarcoma. We found that the pathologic type and time since diagnosis were independent risk factors for the development of SMNs in patients with RMS. The results of this study can help with individualized screening approaches and follow-up timelines of RMS survivors. In the future, we will focus on the survival of RMS cohort and the occurrence of SMN through longer follow-up. The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.
Ethical Approval and Consent to participate Authorization was obtained from the SEER website to collect the data from the database and no additional ethical approval was required.

Consent for publication
Informed consent for publication was collected from all the patients.

Availability of supporting data
The data used and analysed during the current study are available from the SEER program.

Competing interests
The authors have no con icts of interest to declare that are relevant to the content of this article.

Authors' contributions
Author statement: HNZ and FQZ conceptualized and designed the study. ZKL and HG reviewed the literature. JS and JBM collected data and performed statistical analysis. WHW and ZM drafted the manuscript. All authors revised the manuscript. All authors reviewed and approved the nal manuscript.