Clinicopathologic features and outcome of cervical cancer: implications for treatment

Abstract

Objective: We used a regression analysis of the SEER database to establish a new Nomogram for predicting prognosis of cervical cancer patients and guiding the treatment.

Patients and methods: We divided the data into the training cohort and the verification cohort. Univariate and multivariate Cox risk regression analysis was used to identify independent prognostic factors and establish a Nomogram model. The verification cohort was used for external verification, and the accuracy was evaluated with C-index and AUC. Finally, Nomogram was used to establish 1-year, 3-year and 5-year survival curves of cervical cancer patients.

Results: In this study, 5691 patients with cervical squamous cell carcinoma were included. Data obtained from the training cohort were independent risk factors of cervical cancer AJCC stage (p = 0.039), RX Summ - Surgery Primary Site (p = 0.012), radiation (p = 0.031), chemotherapy (p = 0.013), tumor size (p = 0.009), race (p = 0.039). The 1-year, 3-year, and 5-year overall survival rates for cervical cancer patients were 77.2%, 47.8%, and 35.2%, respectively.

Conclusions: The Nomogram model can better screen out more reasonable comprehensive treatments for patients at different stages. And it is of great help to improve the survival rate and reduce the recurrence rate of cervical cancer patients.