Survival prediction in sigmoid-Colon-cancer patients with liver metastasis: a prospective cohort study
Menée à partir de données des registres américains des cancers portant sur 4 981 patients atteints d'un cancer du côlon sigmoïde avec métastases hépatiques (durée médiane de suivi : 20 mois), cette étude identifie des facteurs clinico-pathologiques associés au pronostic et évalue la performance d'un nomogramme intégrant ces facteurs pour prédire la survie spécifique et la survie globale
Purpose : Sigmoid colon cancer (SCC) is a common type of colorectal cancer, frequently leading to liver metastasis. Predicting cause-specific survival (CSS) and overall survival (OS) in SCC with liver metastasis (SCCLM) patients is challenging due to the lack of suitable models.
Methods : Data from SCCLM patients (2010-2017) in the Surveillance, Epidemiology, and End Results (SEER) Program were recruited. Patients were split into training and validation groups (7:3). Prognostic factors were identified using competing risk and Cox proportional hazards models, and nomograms for CSS and OS were developed. Model performance was evaluated with the concordance index and calibration curves, with a two-sided p < .05 was considered statistically significant.
Results : 4,981 SCCLM patients were included, with a median follow-up of 20 months (IQR: 9-33 months). During follow-up, 72.25% of patients died (68.44% from SCC, 3.81% from other causes). Age, race, grade, T stage, N stage, surgery, chemotherapy, CEA, tumor deposits, lung metastasis, and tumor size were prognostic factors for both CSS and OS. The models demonstrated good discrimination and calibration performance, with C-index values of 0.79 (95% CI: 0.78-0.80) for CSS and 0.74 (95% CI: 0.73-0.75) for OS. A web-based application for real-time CSS predictions was created, accessible at https://shuaishao.shinyapps.io/SCCLM/.
Conclusion : Prognostic factors for SCCLM patients were identified basing on SEER database, and nomograms for CSS and OS showed good performance. A web-based application was developed to predict SCCLM-specific survival, aiding in survival risk stratification.
JNCI Cancer Spectrum , article en libre accès, 2023