Risk prediction and treatment assessment in glioma patients using SEER database: a prospective observational study

To use a nomogram to predict the risk of mortality and estimate the impact of current treatment on the prognosis of glioma patients.

A total of 3798 cases were obtained from the Surveillance Epidemiology and End Results database according to the selection criteria. A nomogram was built on the independent clinical factors screened by the variance inflation factor, univariate analyses and a multivariate Cox regression model. Then, categorising the overall population into high-risk, medium-risk and low-risk groups using nomogram-derived risk scores, to study the impact of treatment on different subgroups' survival outcomes. Furthermore, based on the postmatch cohorts, the influences of treatment on survival outcomes were assessed by the log-rank test.

Age, race, stage of disease, histological type, histological grade, surgery, radiotherapy and chemotherapy were identified as the independent prognostic factors. A nomogram with good discrimination and consistency was built. Generally, the patients who underwent surgery, radiotherapy and chemotherapy were more likely to achieve better prognosis than those who did not, except for those who received radiotherapy in the low-risk cohort and those who underwent surgery in the high-risk cohort. Furthermore, the isocitrate dehydrogenase 1/2 (IDH1/2) wild-type patients with surgery, radiotherapy or chemotherapy tended to have higher survival probabilities, while some inconsistent results were observed in the IDH mutant-type cohort.

Surgery, radiotherapy and chemotherapy improved the prognosis, while appropriate selection of topical treatment for the low-risk or high-risk patients deserves further consideration. IDH status gene might be a reliable indicator of therapeutic effectiveness.