4737 - Development and validation of novel nomograms predicting survival of malignant ependymoma patients: A population-based study

Background

Malignant ependymomas are one of the main types of glioma that have high morbidity and mortality. The objective of this study was to construct, validate and calibrate novel nomograms for the survival of ME patients.

Methods

We utilized the Surveillance, Epidemiology, and End Results database (SEER) to retrieve patients with ME. We excluded patients with an unknown diagnosis, no follow up or with autopsy report only. Moreover, we utilized penalized regression models (PRM) by using “hdnom” package in R with the highest AUC and the most stable calibrations to build nomograms for overall (OS) and cancer-specific survival (CSS).

Results

Data of 3435 patients were retrieved and analyzed. We developed 18 PRM and chose Snet and SCAD for OS and CSS. These two models selected age, sex, year of diagnosis, marital status, site, race, type, surgery, radiation, and chemotherapy for the nomograms to be built. Our nomograms had stable calibrations with adequate to good accuracy AUC of 0.70 to 0.76 for the OS nomogram and AUC of 0.75 to 0.78 for the CSS nomogram.

Conclusions

In this population-based study, we developed nomograms that can better help physicians predict the prognosis and survival of ME patients.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Sherief Ghozy.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.