This population-based cohort study suggests that the risk of SMNs is higher in patients with neuroblastoma than in the normal population, with the year of diagnosis affecting the incidence of SMNs. There have been several changes in neuroblastoma treatment since the 1970s, including risk-based stratification of therapy (where the treatment intensity for low- to intermediate-risk patients and high-risk patients is decreased and increased, respectively), autologous transplantation, and MIBG treatment. Update of treatment concept and advancement of treatment technology have led to a significant increase in survival rates, which has allowed observation of long-term side effects represented by SMNs in long-term survivors. Our findings indicate that the time of diagnosis is an independent risk factor for the incidence of SMNs in patients with neuroblastoma. The result showed that patients with a later year of diagnosis are more likely to develop SMN, which may be associated with the evolution of treatment during the period. High-risk patients are recommended to receive higher-intensity radiotherapy and chemotherapy. Chemotherapy-related hematological SMN usually appeared within a few years after treatment. This has led to a significant increase in the risk of total SMN. However, due to the limitation in the number of cases in the SMN cohort, this view still needs more long-term follow-up and a larger sample size for verification.
Applebaum et al. reported that the 10-year cumulative incidence rates of SMNs in high-risk and low-risk patients with neuroblastoma are approximately 1.8% and 0.38%[2], respectively. In a cohort study, Rubino et al. reported a 20-year cumulative incidence of approximately 2.2%[4], which is consistent with the results of the Childhood Cancer Survivor Study (CCSS). Studies have shown that the probability of neuroblastoma survivors developing SMNs greatly increases with an increase in survival time. However, the incidence of SMNs does not tend to plateau over time[1, 5–7]. The results of the above studies indicated that the time to SMN occurrence varied according to the pathological type of SMN. The majority of lymphatic and hematopoietic SMNs occur about 5 years after treatment;sarcoma occurs between 5-15 years after treatment༛and carcinoma usually occurs after 15 years of treatment. With the follow-up time increasing, SMNs are expected to have higher incidence rates.
This study reveals that radiotherapy is not the independent risk factor for developing SMNs, although most studies have indicated the carcinogenic role of radiotherapy in neuroblastoma treatment[4, 7–12]. Compared to other childhood tumors, doses to treat neuroblastoma radiotherapy are low, and low-dose radiation is more likely to cause radiation-related carcinogenesis. However, the aforementioned conclusions may only apply to traditional radiotherapy dominated by external irradiation. We found in univariate analysis that radiotherapy was an independent risk factor for SMNs (p=0.001), but multivariate analysis showed that radiotherapy was not a risk factor for SMNs (p=0.109). 18 of the 38 patients in our SMN cohort had received radiotherapy (47.3%), but we were unable to obtain detailed information such as the field of radiotherapy and the site of the SMNs to assess whether SMN occurred in the radiation field. Another reason that may affect the statistical results is that the SMN cohort has a small sample size, which makes it impossible to calculate statistical differences. Weiss et al.'s cohort showed radiotherapy may not play a role in the occurrence of SMN3,13 mainly because they found that a considerable amount of SMN did not occur in the radiation field, and a small number of SMN patients did not receive radiotherapy; hence, there is insufficient evidence to prove that radiotherapy is implicated in the occurrence of SMN. Another point worth noting is that the wide application of three-dimensional conformal radiotherapy technology (3D-CRT) and intensity-modulated radiotherapy (IMRT) over the past decades has improved the accuracy of treatment and reduced the volume and dose to normal tissues surrounding the radiation field. This may reduce the possibility of SMN occurring around the radiation field. MIBG, which can be broadly considered as a type of radiotherapy, has also been widely applied in neuroblastoma treatment. Weiss et al. argued that MIBG treatment is not an independent risk factor for the incidence of SMNs; moreover, the post-treatment incidence of SMNs does not vary with the increasing treatment times and radiation dose[13]. Similarly, in a cohort study on patients with neuroblastoma, Haupt found that none of the patients with thyroid SMNs had received MIBG treatment[3]. Therefore, these causes of thyroid cancer (SMN) may not be due to MIBG treatment.
Previously, Applebaum et al. suggested that intermediate-risk patients have a significantly higher risk of developing acute myelogenous leukemia even after receiving chemotherapy with low-dose alkylating agents and topoisomerase inhibitors[2]. Danner-Koptik reported a similar conclusion[1, 3, 7]. However, our findings indicate that chemotherapy is not an influencing factor of SMNs, which is inconsistent with other reports[4, 8], and contrary to the current mainstream view. The current mainstream view is that chemotherapy is a risk factor for SMNs, especially cytotoxic drugs represented by alkylating agents and topoisomerase inhibitors, which can cause hematological SMNs such as acute myeloid leukemia (AML). However, there is no clear evidence on whether these drugs can cause SMN for sarcoma and cancer. Applebaum et al.[2] found that SMN also appeared in patients receiving low-dose chemotherapy. Many other studies also showed that SMN caused by chemotherapy is not dose-dependent, which deserves attention[3]. Since the late 1990s, autologous stem cell transplantation has gradually become the standard of care for high-risk neuroblastoma. Studies have shown that the incidence of SMNs in patients with neuroblastoma receiving autologous stem cell transplantation therapy increased from 1.04% at 5 post-transplantation years to 2.6% at 10 post-transplantation years. Moreover, the incidence rate tends to continue to increase with an increasing follow-up duration[7]. Notably, low-risk patients usually only receive surgical therapy; however, they still have a significantly higher probability of developing SMNs than expected, which suggests that genetic factors may be crucially involved in the incidence of SMNs[14].
Additionally, there are many other possible risk factors for developing SMNs. Weiss et al. suggested that patients with relapsed and refractory neuroblastoma are more likely to develop SMNs, which could be associated with genetic factors and higher-intensity chemoradiotherapy[13]. Another example that suggests genes likely playing a role in the development of SMN in patients with neuroblastoma is Federico’s study[1]. He reported that several patients in his SMN cohort had a family history of cancer, but these families did not meet the diagnostic criteria for Li-Fraumeni syndrome (LFS). LFS is a chromosomal dominant genetic disease, which is related to the mutation of the tumor suppressor gene TP53. LFS can cause various cancers, including breast cancer, brain tumors, sarcomas and other cancers. It usually has a clear family history, and most of it occurs at a young age. The probability of aggressive malignant tumors in high-risk family members before the age of 30 is as high as 50%. Osteosarcoma is a common SMN type after childhood tumor treatment, and is related to radiotherapy and chemotherapy but Ewing's sarcoma is not. A patient with Ewing's sarcoma was found in his SMN cohort, which is a rarer type of SMN pathology, meaning that genetic factors, rather than therapeutic factors such as radiotherapy and chemotherapy, may have caused the occurrence of this SMN. Applebaum’s study found that multiple gene mutations exist in SMN patients with neuroblastoma[14], but due to the limitation in the number of cases, no statistical significance was found. The above results all indicate that genetic factors cannot be ignored in the pathogenesis of SMN and need to be further studied.
This study has several limitations. First, the treatment information in the SEER database is incomplete; moreover, there is missing key information, including disease stage, radiotherapy dose and genetic information, which impedes deeper analysis, although our study showed that age, year of diagnosis, degree of pathological differentiation, primary site, chemotherapy, and radiotherapy are independent risk factors for the prognosis of patients with neuroblastoma. We assumed that radiotherapy was performed in patients with high-risk disease, which may be a bias in interpreting the results. Neuroblastoma COG (Children's Oncology Group) risk stratification is mainly based on MYCN gene amplification, age, INSS (International Neuroblastoma Staging System) staging and other factors. Because the SEER database lacks MYCN gene amplification, INSS staging information,. the current data cannot be used for risk stratification. Second, the SEER database does not represent the entire US patient population, which may affect the assessment of the actual incidence of SMNs in patients with neuroblastoma. Furthermore, the patient data included in the database cover a large time span with the risk stratification and treatment strategies of neuroblastoma having significantly changed several times since the 1970s. Therefore, the current stratification information and treatment regimens may not match those of past decades, which could introduce bias into the data. Nevertheless, this study revealed some characteristics regarding the incidence of SMNs in patients with neuroblastoma, which provide support for screening of long-term survivors, as well as early detection and intervention of SMNs.