Establishment and Verification of Prognostic Nomograms for Patients with Gastrointestinal Stromal Tumors: A SEER-Based Study

Autor: Yue-Kui Bai, Yue Zhang, Zhan Chen, Rui-Min Lin
Jazyk: angličtina
Rok vydání: 2019
Předmět:
Zdroj: BioMed Research International, Vol 2019 (2019)
BioMed Research International
ISSN: 2314-6141
2314-6133
Popis: With gastrointestinal tract as the origin, gastrointestinal stromal tumor (GIST) is recognized as the very widespread mesenchymal tumor. A precise prognostic model of survival is required to guide the treatment options of patients with GIST. This study was designed to map the overall survival (OS) and cancer-specific survival (CSS) of GIST patients. According to the Surveillance, Epidemiology, and End Results (SEER) program database, we acquired the data of 6,713 patients with GIST who were diagnosed between 2004 and 2014. We randomly separated the patients into training (n = 4,699) and validation (n = 2,014) groups. To assess the prognostic impact of multiple clinical parameters, the Kaplan-Meier approach and the Cox proportional hazards regression model were adopted, where essential prognostic variables were combined to create nomograms. The consistency index and curve of calibration had been adopted to assess nomogram discrimination ability and prediction accuracy. A multifactor analysis of the training cohort showed that age, gender, size of tumor, location, and primary surgery were remarkably related to survival, and these variables were applied to create nomograms. The nomogram demonstrated excellent accuracy in estimating 2-, 3-, and 5-year OS and CSS, with a C-index of 0.740 (95% confidence interval [CI], 0.723-0.757) for OS and 0.743 (95% CI, 0.718-0.768) for CSS. In the validation cohort, the nomogram-predicted C-index was 0.741 for OS (95%CI, 0.717-0.765) and 0.746 (95%CI, 0.713-0.779) for CSS. All calibration curves showed good consistency between predicted and actual survival. A new nomogram was created and verified to predict the OS and CSS of patients with GIST. These new prognostic models can help enhance the accuracy of survival outcome predictions, thus facilitating to provide constructive therapeutic suggestions.
Databáze: OpenAIRE
Externí odkaz: