The onset characteristics and prognosis of patients with radiation-associated second primary malignancy: a pancancer study in the US SEER cancer registries

Introduction

Second primary malignancies (SPMs) among cancer survivors constitutes a significant proportion of the total cancer incidence in adults, ranging from 11% to 25% (1-4). The occurrence of SPMs has been linked to several critical risk factors, including genetic background, lifestyle, environmental influences, and cancer-related therapies for the first primary malignancy (FPM) (5-10). Radiation therapy (RT) significantly contributes to the development of SPMs (11-13)

Few large population-based studies have comprehensively assessed the risk of developing SPMs associated with RT among cancer survivors, with the majority of these studies only assessing the incidence or risk of SPMs for a single type of FPM or for a limited selection of cancers (14-18). Our previous studies, based on national cancer registries, demonstrated that pelvic RT is associated with an increased risk of secondary bladder cancer, ovarian cancer, and corpus uteri cancer in patients with rectal cancer (19,20). However, additional research is required to determine the influence of RT on SPM occurrence from a pancancer perspective. Furthermore, while recent studies have examined the association between RT and the risk of SPM (2,21-23), none has been performed to identify the human tissues and organs that are prone to developing SPMs after radiation exposure or to identify the onset risk characteristics and prognosis of those with RT associated-SPMs.

In this study, we performed a comprehensive pancancer analysis with the following objectives: (I) evaluate the association between SPM occurrence and radiation exposure in patients with solid FPMs; (II) identify the types of risk-increased SPMs (RI-SPMs) after radiation exposure; and (III) characterize the trends in onset risk and prognosis of patients with RI-SPMs across different types of malignancies, with the aim of strengthening the potential harm of RI-SPMs and facilitating long-term individual surveillance for cancer survivors after RT. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1618/rc).

Methods

Patient selection

Patients diagnosed with solid FPMs were identified from nine registries of the Surveillance, Epidemiology, and End Results (SEER) cancer registries in the period spanning from January 1, 1973, to December 31, 2015. All primary FPM sites were pathologically diagnosed and classified according to the International Classification of Diseases for Oncology, Third Edition (ICD-O-3). All patients with the FPM underwent surgical treatment. The exclusion criteria comprised patients younger than 20 years, those older than 85 years, and individuals with distant-stage cancer. Additionally, patients with missing information regarding age, tumor stage, race, RT, surgery, survival status, or follow-up data were also excluded. For a stable risk calculation, FPM was required to be routinely treated with RT, which we defined as more than 20% of patients undergoing RT as their initial treatment (12). The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Treatment interventions for FPMs

Patients were divided into two groups based on the initial treatment approach for FPMs. Patients in the RT group received RT, and patients in the no RT (NRT) group did not receive RT. The RT types included external-beam RT, radioisotopes, radioactive implants (including brachytherapy), and the combination of beam with implants or isotopes according to the type of malignancies. The SEER program documents data solely for the initial course of cancer treatment. Standard RT doses for each site of SPMs were estimated using established RT protocols (12).

The distance between the border of the RT field and the site of the SPM is an important risk factor that affects SPM occurrence (19,24). To reduce this potential bias, the RT field was classified into head-and-neck radiotherapy (HNRT), thoracic radiotherapy (TRT), and abdominal-and-pelvic radiotherapy (APRT) according to the radiation dose distribution (24). The HNRT field was involved in 12 types of FPMs, including malignancies of the larynx, floor of mouth, gum and mouth, nasopharynx, nasal cavity and middle ear, oropharynx, eye and orbit, hypopharynx, salivary gland, thyroid, tongue, and tonsil. The TRT field was involved in three types of FPMs, including malignancies of the breast, esophagus, and lung and bronchus. The APRT field was involved 10 types of FPMs, including malignancies of the intrahepatic bile duct, stomach, pancreas, rectum and rectosigmoid junction, anal canal and anorectum, anus, corpus uteri, vagina and cervix uteri, prostate, and testis.

Definition, classification, and follow-up of SPMs

The primary outcome was the development of a SPM, defined as any type of SPM occurring more than 5 years after the treatment of the FPM. This time frame was chosen because a minimum latency period of 5 years is required for SPMs to manifest following radiation exposure (25). A total 41 types of solid SPMs were included to enable strong statistical risk estimation. The SEER program distinguishes SPMs from recurrent and metastatic disease of FPMs according to the coding rules of topography of the ICD-O-3 guidelines.

To evaluate the impact of RT on the risk of SPM occurrence, we classified SPMs according to the risk of developing SPMs after RT as follows: risk-increased SPMs (RI-SPMs), risk-decreased SPMs (RD-SPMs), and risk-unaffected SPMs (RU-SPMs). RI-SPMs were those with an increased risk of SPM occurrence after HNRT, TRT, or APRT; RD-SPMs were those with a decreased risk of SPM occurrence after HNRT, TRT, or APRT; and RU-SPMs were those with an unchanged risk of SPM occurrence after HNRT, TRT, and APRT.

Follow-up for SPMs commenced 5 years after the diagnosis of the FPM and continued until the diagnosis of any SPM, the occurrence of all-cause death, or the end of the study period (December 31, 2015), whichever came first. This assumes that the number of SPMs was zero during the first 5 years after RT.

Statistical methods

Multivariable Cox regression analysis was used to evaluate hazard ratio (HR) and 95% confidence interval (CI) of SPMs in patients who underwent RT compared with those who underwent NRT. We calculated the partial regression coefficient (β) of covariates based on the baseline hazard function with multivariable Cox regression analysis. The multivariable Cox regression formula for calculating the β is as follows: h (t, x) = h₀(t) exp (β₁x₁+β₂x₂+β₃x₃+β₄x₄……+β_nx_n). A heatmap was drawn based on the radiotherapy β value in different sites of radiotherapy with hierarchical clustering. The risks were adjusted for potential confounding factors, including race, sex, age at the diagnosis of the FPM, and the calendar year of FPM diagnosis. We conducted a Fine–Gray competing risk regression analysis to further evaluate the cumulative incidence of SPMs, considering non-SPMs and all-cause mortality as competing events. The analyses were performed using R software version 3.5.3 (The R Foundation for Statistical Computing).

The RR was determined using Poisson regression analysis, which provided the relative risk and 95% CIs of SPMs for cancer survivors who received RT compared to those who underwent NRT. Additionally, Poisson regression was employed to calculate the standardized incidence ratio (SIR) and its 95% CIs. The SIR represents the ratio of SPM incidence among survivors to the incidence of malignancies in the general US population. Both RRs and SIRs were adjusted for sex, age at the diagnosis of the FPM, and the calendar year of FPM diagnosis. The excess risk (ER) of SPMs associated with RT was calculated as the number of SPMs in the RT group minus the estimated number of malignancies in the NRT group. RRs were computed using R software version 3.5.3, while SIRs and ERs were determined using SEER*Stat version 8.3.8. To reduce potential surveillance bias in the early follow-up period, analyses of RR, SIR, and ER were restricted to patients who survived 5 years or longer. We evaluated the RRs and SIRs for each type of SPM based on age at FPM diagnosis, latency after FPM diagnosis, and year of FPM diagnosis.

To evaluate the role of RT for FPM in influencing the survival of patients with RI-SPMs, the Kaplan-Meier method was employed to evaluate the 10-year overall survival (OS) for patients with RI-SPMs receiving RT for FPM and patients with RI-SPMs receiving NRT for FPM. The HRs and 95% CIs for 10-year OS were computed using univariate Cox regression analysis, with the P values being determined by the log-rank test. OS was defined as the duration from the diagnosis of RI-SPMs to death from any cause. To reduce potential bias in survival comparisons, propensity score matching (PSM) was employed. All analyses were conducted using R software version 3.5.3.

Results

Patient characteristics

A total of 2,946,359 patients comprising 41 types of solid FPMs were identified in the SEER database (see Table S1). After excluding patients who did not meet our eligibility criteria, 1,406,594 patients comprising 25 types of FPMs were identified as the final cohort. A flowchart of the patient selection is shown in Figure S1. After a latency of 5 years, SPM occurrence was observed in 784,269 patients with FPMs, including 461,516 patients treated with NRT and 322,753 patients treated with RT (Table 1). The median age was 60 years (interquartile range, 51–69 years), and the median follow-up time was 11.5 years (interquartile range, 7.8–16.9 years). The number of patients varied across cancer types, ranging from 442 patients with vaginal malignancy to 301,457 patients with breast malignancy (Table S2). Baseline characteristics of patients with FPMs were compared according to treatment modality (Tables S3-S5).

SPM risk attributable to RT

Multivariable Cox regression analysis was used to evaluate the risk of SPM in patients treated with HNRT, TRT, and APRT and to determine the risk of SPM among survivors who had received RT (Tables S6-S9). Sex, age at FPM diagnosis, calendar year of FPM diagnosis, and administration of RT were included in the multivariable Cox regression analysis, with the main object being the site of the SPM and the latency after the diagnosis of the FPM. The result indicated greater heterogeneity for the risk of SPM occurrence, with HRs ranging from 0.36 (95% CI: 0.25–0.52; β: –1.014) for thyroid malignancy survivors to 4.49 (95% CI: 2.02–9.96; β: –1.502) for nasopharynx malignancy survivors. Statistically significant adjusted β values of RT with multivariable Cox regression analysis of each SPM can be seen in the heatmap in Figure 1.

Figure 1 The risk estimation of SPMs after RT. The β value of RT in each cell was calculated with multivariable Cox regression by dividing the SPMs in patients who received RT by the SPMs in patients who did not receive RT. The β value of RT was adjusted for potential confounding factors, including sex, age at FPM diagnosis, and calendar year of FPM diagnosis. Unsupervised hierarchical clustering of SPMs was completed based on their associations with risk changes after RT, utilizing adjusted β estimates, complete linkage, and an uncentered correlation similarity metric. The vertical axis coordinates of the heatmap are the site of SPMs, and the horizontal axis coordinates are the RT field for FPM. The color scale shows the risk range of β values after RT exposure. Red cells indicate a significantly increased risk of developing SPM after RT (β>0; P<0.05), which was evaluated as the RI-SPM. Green cells indicate a significantly decreased risk of developing SPM after RT (β<0; P<0.05), which was evaluated as the RD-SPM. Yellow cells indicate a nonsignificant risk of developing SPM after RT, which was evaluated as the RU-SPM. The details for β estimates are provided in the supplementary materials. After clustering, the RI-SPMs included 24 types of SPMs, including malignancies of eye and orbit, floor of mouth, tongue, oropharynx, hypopharynx, nasopharynx, larynx, esophagus, lung and bronchus, breast, liver, pancreas, stomach, small intestine, colon, rectum, ovary, corpus uteri, ureter, vagina, urinary bladder, penis, testis, kidney, and renal pelvis; RD-SPMs included 3 types of SPMs, including malignancies of lip, thyroid, and prostate; RU-SPMs included 14 types of SPMs, including malignancies of brain and cranial nerves system, tonsil, nasal cavity and middle ear, gum and mouth, salivary gland, trachea, thymus, intrahepatic bile duct, gallbladder, adrenal gland, vulva, cervix uteri, anus, and anal canal. APRT, abdominal-and-pelvic radiotherapy; NRT, no radiotherapy; TRT, thoracic radiotherapy; HNRT, head-and-neck radiotherapy; RD-SPM, risk-decreased second primary malignancy; RI-SPM, risk-increased second primary malignancy; RU-SPM, risk-unaffected second primary malignancy; SPMs, second primary malignancies; RT, radiation therapy; FPM, first primary malignancy.

To further evaluate the pathogenic factors of SPM occurrence, we classified SPMs into three groups (RI-SPMs, RD-SPMs, and RU-SPMs) based on their associations with risk change after RT, adjusted RT β estimates, complete linkage, and an uncentered correlation similarity metric with unsupervised hierarchical clustering. Analysis of RI-SPMs indicated that 24 types of SPMs that were correlated positively with receiving RT for FPM, including malignancies of the eye and orbit, floor of mouth, tongue, oropharynx, hypopharynx, nasopharynx, larynx, esophagus, lung and bronchus, breast, liver, pancreas, stomach, small intestine, colon, rectum, ovary, corpus uteri, ureter, vagina, urinary bladder, penis, testis, kidney, and renal pelvis. Analysis of the RD-SPMs indicated that three types of SPMs were correlated negatively with receiving RT for FPM, including malignancies of the lip, thyroid, and prostate. Analysis of the RU-SPMs indicated that receiving RT for FPM was not significantly associated with 14 types of SPMs occurrence, including malignancies of the brain and cranial nerves system, tonsil, nose nasal cavity and middle ear, gum and mouth, salivary gland, trachea, thymus, intrahepatic bile duct, gallbladder, adrenal gland, vulva, cervix uteri, anus, and anal canal (Figure 1).

Cumulative incidences of RI-SPMs

To further confirm the association of RT with RI-SPMs, we performed competing risk analysis of the cumulative incidence, comparing patients in the RT group with those in the NRT group. For RI-SPMs, the incidences were 19.8% and 15.3% in the RT and NRT groups, respectively (P<0.001) (Figure 2A). In subgroup analyses of RI-SPMs, the incidences were significantly higher for the RT group than for the NRT group in the HNRT, TRT, and APRT fields (P<0.001) (Figure 2B-2D), which indicated that the development of RI-SPMs was closely associated with receiving RT for an FPM.

Figure 2 Comparisons of cumulative incidence of RI-SPMs between patients treated with RT and those treated with NRT. (A) Cumulative incidences of RI-SPM after RT. (B) Cumulative incidences of RI-SPMs after HNRT. (C) Cumulative incidences of RI-SPMs after TRT. (D) Cumulative incidences of RI-SPMs after APRT. The HRs were adjusted for potential confounding factors, including sex, age at FPM diagnosis, and calendar year of FPM diagnosis. P values were calculated with the log-rank test. RI-SPM, risk-increased second primary malignancy; NRT, no radiation therapy; RT, radiation therapy; HR, hazard ratio; CI, confidence interval; HNRT, head-and-neck radiotherapy; TRT, thoracic radiotherapy; APRT, abdominal-and-pelvic radiotherapy; FPM, first primary malignancy.

Dynamic risk and incidence of RI-SPMs

To thoroughly evaluate the risks for RI-SPMs, we calculated RR values and created dynamic RR plots that incorporated the latency period after the diagnosis of the FPM, the timing of FPM diagnosis, and the age at FPM diagnosis (Tables S10-S13). In patients with RI-SPMs, the risk of SPMs slightly increased in the early latency period at 60–240 months and peaked in the late latency period at 240–360 months (Figure 3A). In the dynamic age-RR plot, the risk of SPM was highest in younger patients and gradually decreased with age (Figure 3B).

Figure 3 Comparison of the dynamic risk and incidence of RI-SPMs. (A) Dynamic RR for RI-SPM in the latency-RR plot. (B) Dynamic RR for RI-SPM in the age-RR plot. (A,B) Adjusted RRs and 95% CIs for the development of RI-SPM in patients treated with RT versus patients treated with NRT are plotted. The detailed data of RRs are provided in the supplementary materials. RR, RT-associated risk; RI-SPM, risk-increased second primary malignancy; HNRT, head-and-neck radiotherapy; TRT, thoracic radiotherapy; APRT, abdominal-and-pelvic radiotherapy; RT, radiation therapy; NRT, no radiation therapy; FPM, first primary malignancy.

We then calculated SIRs and established dynamic SIR plots for patients with FPM treated with RT and for patients with FPM treated with NRT (Tables S14-S17). Among patients with RI-SPMs, the SIRs of the RT group were higher than those of NRT group in the latency-SIR plot, the diagnosis time–SIR plot, and the age-SIR plot. In the latency-SIR plot, the SIRs were stable between 60 and 360 months (Figure S2A-S2D). In the age-SIR plot, an increased incidence was observed in young patients, but a downward trend was observed in older adult patients (Figure S3A-S3D).

In the analyses of each type of RI-SPM, we found that the incidence-related characteristics of SPMs were highly heterogeneous in terms of the latency following FPM diagnosis (Figure S4A) and age at FPM diagnosis (Figure S4B). The details of SIRs in each type of SPM are provided in the Tables S18-S29.

Survival outcome of RI-SPMs

To identify the function of RT for FPM in the survival of patients with RI-SPMs, we compared the 10-year OS between patients with RI-SPMs in the RT group and those in the NRT group. Patients with RI-SPMs receiving RT were matched with patients receiving NRT at a PSM ratio of 1:1 (Tables S30-S33). The results showed that after PSM, the 10-year OS of patients with RI-SPMs who received RT for FPM were significantly shorter than that of those who received NRT for FPM for all sites (HR 1.09, 95% CI: 1.07–1.12; P<0.001) (Figure 4A). The role of RT was the same in the HNRT (HR 1.18; 95% CI: 1.11–1.25; P<0.001), TRT (HR 1.05; 95% CI: 1.02–1.09; P=0.002), and APRT fields (HR 1.12; 95% CI: 1.08–1.15; P<0.001) (Figure 4B-4D).

Figure 4 Comparison of survival outcomes among those with RI-SPMs. (A) Survival comparison between patients with RI-SPM receiving RT for FPM and patients receiving no NRT for FPM in all cancer sites (after PSM). (B) Survival comparison between patients with RI-SPM receiving HNRT for FPM and patients receiving NRT for FPM in head-and-neck cancer (after PSM). (C) Survival comparison between patients with RI-SPM receiving TRT for FPM and patients receiving NRT for FPM in thoracic cancer (after PSM). (D) Survival comparison between patients with RI-SPM receiving APRT for FPM and patients receiving NRT for FPM in abdominal and pelvic cancer (after PSM). (A-D) Patients who developed RI-SPMs and received RT for FPMs in all sites were matched with patients who did not receive RT for first FPMs in all sites at a PSM ratio of 1:1. The matched variables for PSM included age at RI-SPM diagnosis, year of RI-SPM diagnosis, sex, grade of RI-SPM, stage of RI-SPM, surgery for RI-SPM, chemotherapy for RI-SPM, and RT for RI-SPM. The detailed patient characteristics of patients who developed RI-SPMs before and after PSM are shown in the supplementary data. HRs were calculated using Cox regression. RI-SPM, risk-increased second primary malignancy; NRT, no radiation therapy; RT, radiation therapy; HR, hazard ratio; CI, confidence interval; HNRT, head-and-neck radiotherapy; TRT, thoracic radiotherapy; APRT, abdominal-and-pelvic radiotherapy; PSM, propensity score matching.

Discussion

Recent population-based incidence data indicate that cancer survivors have higher risks of developing SPMs than do the general population (16,26-30). We conducted the first large-scale, population-based pancancer study to identify RI-SPMs based on the risk of developing SPMs following radiation exposure. We also examined the onset characteristics and long-term prognosis of RI-SPMs after RT among cancer survivors. Our main findings are as follows. First, RT was found to be associated with an increased overall risk of SPMs, but there was substantial heterogeneity regarding the RT-associated cancer risk in different types of SPMs. Second, 24 types of RI-SPMs were identified to further evaluate the impact of RT on the risk of developing SPMs. Third, the incidence of RI-SPMs was higher in survivors after RT than in those undergoing NRT. Fourth, the onset risk of RI-SPMs decreased with age and increased with latency, but the risk dynamics varied across different types of RI-SPMs. Fifth, patients with RI-SPMs who underwent RT had a worse prognosis compared with matched patients undergoing NRT.

Previous studies have explored the risk of RT-associated SPMs, but the majority examined a single cancer type or single RT model in determining the risk of SPM occurrence, which involves small sample sizes and large selection biases (31,32). In our study, we evaluated the risk of SPMs after RT using a large sample size and long-term follow-up. The advantage of using the SEER database is that its systematic coding rules for SPM have not markedly changed over time. Furthermore, our findings are more in line with the actual clinical situation and have more significance for clinical guidance in SPM occurrence after RT. Moreover, all types of solid FPMs were selected as the first cancer to evaluate the risk of individual types of SPMs. We also considered the effect of RT fields on SPM occurrence, and these measures cumulatively could limit the bias arising from small sample sizes and RT doses. In addition, our analysis included data spanning 30 years, allowing for a comprehensive examination of the incidence and characteristics of SPMs after RT.

We combined several statistical methods to systematically evaluate the association between SPMs and RT. We first classified SPMs using Cox regression and used competitive risk regression to confirm the results. We use the SIR to compare the incidence of SPMs in cancer survivors with that in the general US population, providing a more comprehensive understanding of the risk and characteristics associated with SPM incidence. The choice of latency cutoff points significantly affects the risk assessment for the development of SPMs following RT, which may account for the discrepancies observed in the risk of SPM occurrence reported by previous studies. We identified SPMs that developed after a 5-year latency from radiation exposure for solid tumors because this cutoff point enabled us to draw a more reasonable conclusion (12,25).

The dynamic risk and incidence of SPMs after RT are highly heterogeneous across different cancer types. Overall, the RT-associated risk for developing SPMs was significantly higher in patients with FPMs after RT than in the general population. In most latency periods, there was no significant change in the risk of developing SPMs after RT. These findings underscore the importance of long-term follow-up for the early detection of SPMs after RT following malignancies of the nasopharynx, hypopharynx, and vagina, but follow-up for malignancies of tongue and floor of mouth should also be considered in the early latency period. Furthermore, while there was a general trend toward a decreased risk of RI-SPMs with increasing age at diagnosis, and this trend varied significantly across different cancer types. This suggests that younger patients who undergo RT are at a higher risk of developing SPMs and should therefore be closely monitored during follow-up.

Compared to FPMs, the onset characteristics and survival outcomes of RI-SPMs may exhibit considerable heterogeneity. We found that patients with RI-SPMs experienced worse prognoses compared to matched controls treated with NRT, which is consistent with our previous results (19,20). The poorer prognosis of those with RI-SPMs may be attributed to the involvement of different genetic signaling pathways triggered by radiotherapy, which differ from those associated with primary malignancies. Additionally, due to their distinct tumorigenesis, RI-SPMs tend to develop resistance to conventional treatment methods, which further contributes to a poor prognosis.

This study involved several limitations which should be acknowledged. First, due to the observational nature of the study, the initial treatment for FPMs was not randomized, which might have introduced potential confounding factors. These include, but are not limited to, mortality from the FPM, effects of prior treatments such as chemotherapy, genomic susceptibility to cancers, and exogenous factors like smoking history for tobacco-related cancers. These factors could influence both the risk of cancers commonly treated with radiotherapy (such as head, neck, and lung cancers) and overall mortality (2,16,18,21,22,26,28,33-35). Therefore, it is challenging to balance this array of influencing factors between the treatment groups. To address this, we employed multiple approaches, including multivariable Cox models and competing risk models, to adjust for these and other confounding variables and mitigate potential bias. Second, the impact of confounding factors may vary across different types of FPMs, as factors influencing the administration of RT can differ by cancer site. This variability could lead to either an overestimation or underestimation of RT-related risks. To minimize site-related bias, we classified FPMs according to the RT field. However, it is important to note that only the general location of the radiation field is known, not the specific field borders. Additionally, data on dosimetric exposure to tissues within the region of SPM—whether the SPM occurred within the radiation field, near its edge, or remotely—were not available. This limitation could potentially affect the accuracy of the risk assessment. Third, the SEER database records only the initial treatment information for FPMs. Information on whether patients received delayed RT in subsequent treatments is not available, and this could have potentially led to misclassification of patients who received RT into the NRT group. Although this is unlikely to have altered our primary conclusions, there might have been an underestimation of the increased risk associated with RT. Fourth, the dose-response relationship between RT and the risk of SPMs could not be assessed due to the lack of information on RT doses in the SEER database. Fifth, the absence of available genomic data limits our ability to explain the poor prognosis of those with RI-SPMs following radiotherapy exposure.

Conclusions

This study is the first to identify the types of RI-SPMs and to identify the onset characteristics and prognoses of patients RI-SPMs and may thus help inform more targeted individualized follow-up for patients treated with RT. Our findings regarding the onset characteristics and prognosis of SPM survivors after RT provide a valuable theoretical basis for further ascertaining the impact of RT on SPMs.

Acknowledgments

The authors thank Fahad Mukhtar MD, MPH (Department of Epidemiology and Biostatistics, College of Public Health, University of South Florida), for professional comments and valuable suggestions that have enabled us to further improve our manuscript.

Funding: This work was supported by the Shanxi Provincial Key Medical Research Project (No. 2020XM5) and the Science and Education Cultivation Fund of the National Cancer and Regional Medical Centre of Shanxi Provincial Cancer Hospital (Nos. TD2023001 and SD2023034).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1618/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1618/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1618/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Hyun N, Leach CR, et al. Association of First Primary Cancer With Risk of Subsequent Primary Cancer Among Survivors of Adult-Onset Cancers in the United States. JAMA 2020;324:2521-35. [Crossref] [PubMed]
2. Liu J, Zheng Q, Beeraka NM, et al. Long-Term Risk of Subsequent Malignant Neoplasms Among Childhood and Adolescent Lymphoma Survivors (1975-2013): A Population-Based Predictive Nomogram. Oncologist 2023;28:e765-73. [Crossref] [PubMed]
3. Murphy CC, Gerber DE, Pruitt SL. Prevalence of Prior Cancer Among Persons Newly Diagnosed With Cancer: An Initial Report From the Surveillance, Epidemiology, and End Results Program. JAMA Oncol 2018;4:832-6. [Crossref] [PubMed]
4. Zhao X, Ji J, Li Y, et al. The risk and survival of multiple myeloma as the second primary malignancy in a single Chinese center. Transl Cancer Res 2024;13:2905-12. [Crossref] [PubMed]
5. Travis LB, Demark Wahnefried W, Allan JM, et al. Aetiology, genetics and prevention of secondary neoplasms in adult cancer survivors. Nat Rev Clin Oncol 2013;10:289-301. [Crossref] [PubMed]
6. Yang J, Gao J, Hu J, et al. Carbon-ion radiotherapy in the treatment of radiation-induced second primary malignancies. Ann Transl Med 2022;10:1200. [Crossref] [PubMed]
7. Fillon M. Genetic factors contribute to subsequent neoplasms in survivors of childhood cancer. CA Cancer J Clin 2020;70:143-4. [Crossref] [PubMed]
8. Reulen RC, Frobisher C, Winter DL, et al. Long-term risks of subsequent primary neoplasms among survivors of childhood cancer. JAMA 2011;305:2311-9. [Crossref] [PubMed]
9. Mohamad O, Tabuchi T, Nitta Y, et al. Risk of subsequent primary cancers after carbon ion radiotherapy, photon radiotherapy, or surgery for localised prostate cancer: a propensity score-weighted, retrospective, cohort study. Lancet Oncol 2019;20:674-85. [Crossref] [PubMed]
10. Turcotte LM, Neglia JP, Reulen RC, et al. Risk, Risk Factors, and Surveillance of Subsequent Malignant Neoplasms in Survivors of Childhood Cancer: A Review. J Clin Oncol 2018;36:2145-52. [Crossref] [PubMed]
11. Mariscotti G, Durando M, Ghione G, et al. Breast cancer surveillance in patients treated by radiotherapy for Hodgkin's lymphoma. Radiol Med 2013;118:401-14. [Crossref] [PubMed]
12. Han C, Wu Y, Kang K, et al. Long-term radiation therapy-related risk of second primary malignancies in patients with lung cancer. J Thorac Dis 2021;13:5863-74. [Crossref] [PubMed]
13. Berrington de Gonzalez A, Curtis RE, Kry SF, et al. Proportion of second cancers attributable to radiotherapy treatment in adults: a cohort study in the US SEER cancer registries. Lancet Oncol 2011;12:353-60. [Crossref] [PubMed]
14. Moschini M, Zaffuto E, Karakiewicz PI, et al. External Beam Radiotherapy Increases the Risk of Bladder Cancer When Compared with Radical Prostatectomy in Patients Affected by Prostate Cancer: A Population-based Analysis. Eur Urol 2019;75:319-28. [Crossref] [PubMed]
15. Rombouts AJM, Hugen N, Elferink MAG, et al. Incidence of second tumors after treatment with or without radiation for rectal cancer. Ann Oncol 2017;28:535-40. [Crossref] [PubMed]
16. Leng J, Qiu H, Huang Q, et al. Recommendations for broadening eligibility criteria in esophagus cancer clinical trials: the mortality disparity of esophagus cancer as a first or second primary malignancy. J Thorac Dis 2024;16:3882-96. [Crossref] [PubMed]
17. Sunguc C, Hawkins MM, Winter DL, et al. Risk of subsequent primary oral cancer in a cohort of 69,460 5-year survivors of childhood and adolescent cancer in Europe: the PanCareSurFup study. Br J Cancer 2023;128:80-90. [Crossref] [PubMed]
18. Geurts YM, Neppelenbroek SIM, Aleman BMP, et al. Treatment-specific risk of subsequent malignant neoplasms in five-year survivors of diffuse large B-cell lymphoma. ESMO Open 2024;9:102248. [Crossref] [PubMed]
19. Guan X, Wei R, Yang R, et al. Association of Radiotherapy for Rectal Cancer and Second Gynecological Malignant Neoplasms. JAMA Netw Open 2021;4:e2031661. [Crossref] [PubMed]
20. Guan X, Wei R, Yang R, et al. Risk and Prognosis of Secondary Bladder Cancer After Radiation Therapy for Rectal Cancer: A Large Population-Based Cohort Study. Front Oncol 2021;10:586401. [Crossref] [PubMed]
21. Feng Y, Qian K, Guo K, et al. Effectiveness and risk of second primary malignancies after radiotherapy in major salivary gland carcinomas: A retrospective study using SEER database. Head Neck 2024;46:1201-9. [Crossref] [PubMed]
22. Smith J, Cust AE, Lo SN. Risk factors for subsequent primary melanoma in patients with previous melanoma: a systematic review and meta-analysis. Br J Dermatol 2024;190:174-83. [Crossref] [PubMed]
23. Wallis CJ, Mahar AL, Choo R, et al. Second malignancies after radiotherapy for prostate cancer: systematic review and meta-analysis. BMJ 2016;352:i851. [Crossref] [PubMed]
24. Stovall M, Weathers R, Kasper C, et al. Dose reconstruction for therapeutic and diagnostic radiation exposures: use in epidemiological studies. Radiat Res 2006;166:141-57. [Crossref] [PubMed]
25. Preston DL, Ron E, Tokuoka S, et al. Solid cancer incidence in atomic bomb survivors: 1958-1998. Radiat Res 2007;168:1-64. [Crossref] [PubMed]
26. Lu J, Chen D, Shen Z, et al. Impact of radiotherapy on second primary lung cancer incidence and survival in esophageal cancer survivors. Sci Rep 2024;14:17720. [Crossref] [PubMed]
27. Eid E, Maloney NJ, Cai ZR, et al. Risk of Multiple Primary Cancers in Patients With Merkel Cell Carcinoma: A SEER-Based Analysis. JAMA Dermatol 2023;159:1248-52. [Crossref] [PubMed]
28. Wan M, Wu J, Jiang Z, et al. Risk of second primary cancers in patients with rectal neuroendocrine neoplasms: a surveillance, epidemiology, and end results analysis. Front Oncol 2023;13:1248268. [Crossref] [PubMed]
29. Yang Z, Liu L, Leng K, et al. Risk of second primary malignancies in survivors of pancreatic neuroendocrine neoplasms from 2000 to 2018. J Gastroenterol Hepatol 2023;38:1474-84. [Crossref] [PubMed]
30. Wood ME, Vogel V, Ng A, et al. Second malignant neoplasms: assessment and strategies for risk reduction. J Clin Oncol 2012;30:3734-45. [Crossref] [PubMed]
31. Lee JM, Buist DS, Houssami N, et al. Five-year risk of interval-invasive second breast cancer. J Natl Cancer Inst 2015;107:djv109. [Crossref] [PubMed]
32. Chaturvedi AK, Engels EA, Gilbert ES, et al. Second cancers among 104,760 survivors of cervical cancer: evaluation of long-term risk. J Natl Cancer Inst 2007;99:1634-43. [Crossref] [PubMed]
33. Casey DL, Vogelius IR, Brodin NP, et al. Risk of Subsequent Neoplasms in Childhood Cancer Survivors After Radiation Therapy: A PENTEC Comprehensive Review. Int J Radiat Oncol Biol Phys 2024;119:640-54. [Crossref] [PubMed]
34. Kim MS, Lee SJ, Lee MH, et al. Risk of Subsequent Primary Cancer in Thyroid Cancer Survivors: A Nationwide Population-Based Study. Diagnostics (Basel) 2023;13:2903. [Crossref] [PubMed]
35. Choi YY, Lee M, Kim EH, et al. Risk of Subsequent Primary Cancers Among Adult-Onset 5-Year Cancer Survivors in South Korea: Retrospective Cohort Study. JMIR Public Health Surveill 2024;10:e48380. [Crossref] [PubMed]