New nomograms to predict overall and cancer‐specific survival of angiosarcoma
Objective This study was designed to establish and validate promising and reliable nomograms for predicting the survival of angiosarcoma (AS) patients. Methods The Surveillance, Epidemiology, and End Results database was queried to collect the clinical information of 785 AS patients between 2004 and...
Saved in:
Published in | Cancer medicine (Malden, MA) Vol. 11; no. 1; pp. 74 - 85 |
---|---|
Main Authors | , , , , , |
Format | Journal Article |
Language | English |
Published |
United States
John Wiley & Sons, Inc
01.01.2022
John Wiley and Sons Inc Wiley |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Objective
This study was designed to establish and validate promising and reliable nomograms for predicting the survival of angiosarcoma (AS) patients.
Methods
The Surveillance, Epidemiology, and End Results database was queried to collect the clinical information of 785 AS patients between 2004 and 2015. Data were split into a training cohort (n = 549) and a validation cohort (n = 236) without any preference. Univariate Cox and multivariate Cox regression analyses were performed to analyze the clinical parameters. Independent prognostic factors were then identified. Two nomograms were constructed to predict overall survival (OS) and cancer‐specific survival (CSS) at 3 and 5 years. Finally, the models were evaluated using concordance indices (C‐indices), calibration plots, and decision curve analysis (DCA).
Results
Based on the inclusion and exclusion criteria, 785 individuals were included in this analysis. Univariate and multivariate Cox regression analyses revealed that age, tumor size, and stage were prognostic factors independently associated with the OS of AS. Tumor site, tumor size, and stage were associated with the CSS of AS. Based on the statistical results and clinical significance of variables, nomograms were built. The nomograms for OS and CSS had C‐indices of 0.666 and 0.654, respectively. The calibration curves showed good agreement between the predictive values and the actual values. DCA also indicated that the nomograms were clinically useful.
Conclusion
We established nomograms with good predictive ability that could provide clinicians with better predictions about the clinical outcomes of AS patients.
We constructed new and reliable nomograms with the clinical information obtained from the Surveillance, Epidemiology, and End Results database. Internal and external validation results suggested that our models had good prediction effects. |
---|---|
Bibliography: | Yuan‐Yuan Liu and Bu‐Shu Xu should be considered joint first author. ObjectType-Article-2 SourceType-Scholarly Journals-1 ObjectType-Undefined-1 ObjectType-Feature-3 content type line 23 |
ISSN: | 2045-7634 2045-7634 |
DOI: | 10.1002/cam4.4425 |