Epidemiological Characteristics and Prognosis Model of Pineal Region Tumors: A Retrospective Analysis Based on SEER Database.

BACKGROUND: Pineal region tumor is a rare intracranial tumor, and its special location leads to its own characteristics. This study aimed to provide some insight to the medical practice for the care of pineal region tumor. It investigated key epidemiological characteristics and survival prognosis of pineal tumors based the epidemic data from SEER database.

METHODS: Data of pineal region tumor patients from 1975 to 2019 were extracted from the SEER database and divided all data into three pathologic groups: "Germ cell tumors (GCTs)", "Pineal parenchymal tumors (PPTs) " and "Others". Patients' overall survival (OS) was analyzed using the Kaplan-Meier (K-M) method. Prognostic effect of patient's characteristics on OS was explored via the Cox proportional hazard model. Analyses results were presented in tables and figures including K-M plots, forest plot and nomogram. A calibration curve was used to verify that the nomogram. All analyses were performed for all patients overall and by pathological group using SPSS and R language.

RESULTS: Based on pre-defined the inclusion and exclusion criteria, 628 patients were included in this study, 440 (70.1%) males and 188 (29.9%) females. The majority of patients was between 0-19 years; 225 (35.8%) of patients' pathological type was germinoma. Age, surgery, behavioral codes and pathology were significant factors of OS. A calibration curve was used to verify that the nomogram had a good prediction effect.

CONCLUSION: An intuitive nomogram was developed and verified. It can predict the prognosis of patients with pineal tumors.