Abstract
Background
Current staging system classifies all patients with metastatic cervical cancer as stage IVB, without accounting for prognostic heterogeneity. This study aimed to develop an exploratory metastatic pattern-based framework for prognostic stratification within stage IVB cervical cancer.
Methods
Patients with
de novo
stage IVB cervical cancer were identified from the Surveillance, Epidemiology, and End Results database and the Cancer Hospital of Shantou University Medical College. Random survival forest and Cox regression were used to identify key predictors of cancer-specific survival and develop a nomogram-derived scoring system for stratifying patients into IVB subgroups. Prognostic differences were evaluated using Kaplan-Meier and Cox regression analyses. Exploratory analyses further examined survival patterns according to treatment modality across subgroups.
Results
Brain, liver, lung, and bone metastases were identified as the main metastatic determinants of prognosis. Based on their prognostic contributions and combinations, patients were categorized into three subgroups: stage IVB1 (metastases excluding brain, liver, lung, or bone), stage IVB2 (metastases excluding brain, including one of liver, lung, or bone, or combinations of liver and bone, or lung and bone), and stage IVB3 (metastases to the brain, or concurrent liver and lung). The three groups showed distinct survival patterns in the training and internal validation cohorts. In the external cohort, the overall prognostic ordering was similar; however, validation of the IVB3 category remained exploratory owing to its limited representation. In exploratory subgroup analyses, chemoradiotherapy was associated with longer survival than chemotherapy alone in stage IVB1 and stage IVB2, whereas no significant survival difference was observed between these approaches in stage IVB3.
Conclusion
The proposed IVB1-IVB3 classification represents an investigational prognostic framework for characterizing heterogeneity within stage IVB cervical cancer. Treatment-related findings were observational and hypothesis-generating and should not be interpreted as evidence of treatment benefit. Prospective multicenter studies are required before this framework can be considered for clinical adoption.