Abstract
OBJECTIVES: To study the risk factors for the onset of secondary primary malignancies (SPM) and the latency between SPM and cervical cancer after radiotherapy. METHODS: We selected patients with cervical cancer who underwent radiotherapy between 2000 and 2019 from the Surveillance, Epidemiology, and End Results (SEER) database. And the data of patients with cervical cancer who underwent radiotherapy in Guangxi Medical University Cancer Hospital during January 1,1997 to December 31,2016 were collected and analyzed. The factors associated with SPM onset and latency were then estimated by nomograms based on logistic regression and a complete risk model. Dynamic risk plots were performed by Poisson regression. RESULTS: A total of 32,313 cases of cervical cancer who underwent radiotherapy were downloaded from the SEER database; of these, 19,439 cases had a complete dataset and were included in the final analysis. In total, 561 cases suffered from SPM; the remaining 18,878 did not. And a total of 1486 cases of cervical cancer who underwent radiotherapy from Guangxi Medical University Cancer Hospital were analyzed, 27 cases caught SPM and the rest of 1459 cases did not. Patients with SPM were older than those without SPM(p=0.000); significant associations were also identified between SPM and white race(p=0.000), localized stage (p=0.000), squamous carcinoma (SCC)(p=0.003), surgery(p=0.000), and combination radiotherapy (p=0.026). A logistic regression nomogram showed that older age (HR:1.015, 95%CI:1.009-1.021, p=0.000), localized stage (HR:4.056, 95%CI: 2.625-6.269, p=0.000) and regional stage (HR: 3.181, 95%CI:2.094-4.834, p=0.000), white (HR: 1.722, 95%CI:1.145-2.590, p=0.000) and black race (HR: 1.889, 95%CI:1.327-2.689, p=0.000), and the receipt of surgery (HR: 1.381, 95%CI:1.151-1.657, p=0.000) were all independent risk factors for the onset of SPM. The largest proportion of cases involved SPM in the female reproductive system. A dynamic risk plot showed that age, race, stage, and surgery had impacts on the latency of SPM onset. A competing risk regression analysis nomogram showed that age (HR: 1.564, 95%CI: 1.272-1.920, p=0.000), surgery (HR: 1.415, 95%CI: 1.140-1.760, p=0.002), localized stage (HR: 8.035, 95%CI: 4.502-14.340, p=0.000) and regional stage (HR: 4.904, 95%CI: 2.790-8.620, p=0.000), and black race (HR: 1.786, 95%CI: 1.161-2.750, p=0.008) all had significant impacts on the cumulative incidence and latency of SPM. CONCLUSIONS: Advanced age, the receipt of surgery, earlier stages, and white and black race were identified as risk factors for SPM onset and influenced latency in patients with cervical cancer after radiotherapy.