Prognostic factors and nomogram prediction of survival probability in primary spinal cord astrocytoma patients

OBJECTIVE Knowledge on the management of spinal cord astrocytoma (SCA) remains scarce. Here, the authors constructed and validated a predictive nomogram, often used for individualized prognosis and evaluations, to estimate cancer-specific survival (CSS) and overall survival (OS) for patients with SCA. METHODS Epidemiological characteristics were compared between low-grade SCA (LGSCA) and high-grade SCA (HGSCA) patients from the Surveillance, Epidemiology, and End Results (SEER) database. Risk factors for CSS and OS were determined using univariate and multivariate analyses and Kaplan-Meier curves. A nomogram was developed to individually predict the 3-, 5-, and 10-year CSS and OS rates. The clinical usefulness of the nomogram was assessed using calibration plots, the concordance index (C-index), and time-dependent receiver operating characteristic curves. RESULTS A total of 468 LGSCA and 165 HGSCA patients were eligible for inclusion. LGSCA and HGSCA patients demonstrated differences in age, tumor extension, insurance status, adjuvant treatment, and survival. Multivariate analysis demonstrated that in the LGSCA group, tumor extension, surgery type, and adjuvant therapy were individually associated with CSS. The distance of tumor extension and WHO grade were individually associated with CSS in the HGSCA group. The prognostic variables were further demonstrated using the Kaplan-Meier method, which also suggested that adjuvant treatment provided no advantage to HGSCA patients. A nomogram was constructed, and the C-index for CSS was 0.84 by internal validation (95% CI 0.79-0.90). CONCLUSIONS This research suggests that the distance of tumor extension, type of surgery, and adjuvant therapy are significant risk factors for CSS using multivariate analysis in the LGSCA group. Adjuvant treatment provided no advan- tages for CSS or OS in patients with HGSCAs. The nomogram may be clinically useful to healthcare providers.