Survival beyond treatment: patterns and determinants of disease-free survival in cervical cancer at a tertiary care center in India

Abstract

Introduction Cervical cancer remains a major public health concern in India, with a substantial burden of late-stage diagnosis and limited survival data from tertiary care settings. Disease-free survival (DFS) is a key indicator of treatment effectiveness and recurrence risk, yet evidence on its determinants in Indian populations remains limited. This study aimed to assess DFS pattern and identify prognostic factors among cervical cancer patients treated at a tertiary care center in India. Data and methods A retrospective–prospective cohort study was conducted at Sir Sunderlal Hospital, Banaras Hindu University, Varanasi. The study included 587 cervical cancer patients diagnosed between 2011 and 2021, with follow-up until February 2025. Disease-free survival was defined as the time from first follow-up after treatment completion to recurrence. Kaplan–Meier methods, log-rank tests, and Cox proportional hazards models with time-varying covariates were used to estimate survival and identify predictors. Results The median follow-up duration was 23.4 months (IQR: 10.2–52.6). Most patients were aged ≤50 years (51.8%), from rural areas (68.5%), and presented with advanced disease (FIGO stage II: 36.3%; stage III: 45.8%). Squamous cell carcinoma accounted for 91.8% of cases, and 76.0% had tumor size ≥4 cm. Kaplan–Meier analysis showed a sharp decline in DFS during early follow-up, indicating a higher risk of recurrence in the initial years. Significant differences in DFS were observed by age (p=0.002), tumor size (p=0.005), vaginal involvement (p=0.011), lymph node involvement (p=0.012), and addiction status (p=0.023). In multivariable analysis, age ≤50 years (AHR = 1.64; 95% CI: 1.16–2.34), tumor size ≥4 cm (AHR = 2.98; 95% CI: 1.44–6.15), and vaginal involvement (AHR = 1.56; 95% CI: 1.06–2.29) were independently associated with poorer DFS. Tumor size also showed a significant time-varying effect (HR = 0.97, p=0.010). Conclusion DFS in cervical cancer is strongly influenced by tumor burden and locoregional spread. The early decline in survival highlights the need for intensive follow-up during the first two years post-treatment. Targeted interventions focusing on early detection and risk-stratified management are essential to improve outcomes in resource-limited settings.