Radiotherapy is a crucial component of the comprehensive treatment for GC, contributing significantly to overall treatment outcomes[6, 24]. However, the potential toxic effects of RT on patients should not be ignored, especially the radiation exposure associated with RT may lead to the development of SPMs, an easily overlooked complication. Despite advancements in RT techniques mitigating inadvertent irradiation of adjacent normal tissues, some studies[15, 25, 26] indicate an increased risk of SPMs after RT for specific cancers. In this study, leveraging data from a substantial population-based cohort, we employed various methods to unravel the intricate relationship between RT and individual SPMs risk in resectable GC patients, simultaneously evaluating the prognostic impact of RT on primary GC patients. Our findings indicate that RT not only does not increase the risk of developing second primary hematologic or solid malignancies in GC patients but also improves OS and DSS in these individuals.
RT is widely acknowledged as a significant risk factor for specific SPMs across various primary cancer types. For instance, in a study by Guan et al[13], involving 20,142 female patients with rectal cancer as the first primary cancer who survived for at least 5 years, various ways of statistical analysis revealed a significant increase in the risk of uterine corpus and ovarian cancer following RT. On a different note, Zhou et al[20]. analyzed 62,435 patients with nonmetastatic lung cancer who underwent surgery, demonstrating a significant association between RT and an increased risk of major second primary malignant tumors (RR = 1.21; 95% CI, 1.08–1.35), particularly in the development of second primary gastrointestinal malignancies (RR = 1,77, 95%CI: 1.44–2.15). Conversely, there was no association with the development of second hematologic malignancies (RR = 1.08, 95% CI: 0.84–1.37).
In the context of the SPMs in GC, Chen et al[27]. identified RT as an independent risk factor for SPMs in a study involving 137,798 participants from the Taiwan National Health Insurance database. Binay et al[28]. conducted an evaluation of 33,720 patients drawn from the SEER 13 database, employing standardized incidence rate ratios (SIRs). Their research has demonstrated a significantly elevated risk of developing SPMs among GC patients in the United States, when compared to the general population. Additionally, the research highlighted a noteworthy increase in SPMs risk during RT for GC. Similarly, Jin et al[29]. analyzed data on patients diagnosed with gastric adenocarcinoma (GAC) in the SEER database from 2000 to 2020. The results showed that patients with GAC had a significantly increased risk of developing SPMs, with a SIRs of 1.36, 95% CI of 1.32–1.40 and an excess incidence risk (EAR) of 53.57. Furthermore, RT was associated with an increased risk of developing SPMs following a diagnosis of GAC. Nonetheless, regarding the impact of RT on SPMs in GC, the previous studies possessed certain limitations. Firstly, it did not account for the latency period of radiation exposure. Secondly, it omitted the consideration of the influence of early death on the occurrence of SPMs.
RT does not significantly increase the likelihood of specific SPMs across all primary malignancies, and there have been instances of conflicting results from the same site in previous studies. To improve the scientific validity and rigor of this study, we drew on the experience of RT for other malignancies than stomach and analyzed the reasons for these discrepancies, including factors such as the definition of the SPMs in each study, the latency period chosen, the follow-up time, the statistical methods used, and the sample size of the cohort population. Given the relatively low prevalence of SPMs, we utilized data from 17 registries of the SEER database, incorporating a broader population to increase the confidence of our results. This difficulty arises from the challenge of distinguishing between primary and recurrent tumors in the residual stomach within the SEER database. Therefore, we opted to exclude patients with a second tumor in the remnant stomach. To reduce bias in this retrospective study and improve the analysis's robustness, we employed a 1:1 PSM method to pair populations according to predefined inclusion and exclusion criteria. Additionally, recognizing variations in the latency periods for radiation-induced hematological and solid tumors, we established two cohorts: one comprising patients with a minimum survival of 2 years and the other with a minimum survival of 5 years, aligning with the briefest latency period for the development of second hematological and solid tumors, respectively. The choice of statistical analysis methods has a significant impact on the interpretation of the results. We applied a variety of statistical methods, including Poisson regression and competing risks regression, to fully validate the results of the study. The results showed that both methods consistently indicated that RT does not increase the risk of secondary hematologic and solid malignancies. In addition, patients were categorized into subgroups based on clinical characteristics with the aim of accurately assessing potential RT risk.
Our study has several limitations. First, being a population-based study, potential bias may arise due to the non-randomized allocation of the initial treatment, leading to an inability to eliminate bias resulting from imbalances between patients receiving and not receiving RT. Second, the absence of detailed data on postoperative RT modalities, dose, and frequency limited our exploration of the dose-response relationship between RT and SPMs development. Third, the occurrence of SPMs may not solely be associated with RT exposure but could also be influenced by other significant risk factors, such as lifestyle, genetic background, environmental factors, and other cancer-related treatments, which cloud be further elucidated in prospective cohort study or clinical trials.