Trends In The Incidence And Analysis Of Prognostic Factors Of Adult-onset Medulloblastoma

Purpose:Medulloblastoma(MB)only accounts for 1% of all central nervous system tumors in adults.Due to limited population-based evidence,it remains challenging to comprehensively evaluate the epidemiological trends and to define criteria for risk stratification in the investigation of adult MB.This study aimed to explore populationlevel trends in the incidence of adult MB and to examine potential predictors for overall survival(OS).Patients and methods:Data for patients from 1973 to 2015 were extracted from the Surveillance,Epidemiology and End Results(SEER)database.Age-adjusted Incidence Rates(AAIR)and Relative Survival Rate(RSR)were calculated for groups divided on the basis of age,gender,and ethnicity.Propensity-score matching(PSM)was used to compensate for potential selection biases.Cox proportional hazard models were used to examine potential predictors for OS.The Nomogram was established which was based on the multivariable analysis to predict the OS.Calibration plots and Harrell’s concordance index(C-index)were used to evaluate the performance of the Nomogram model.Results:AAIR was stable among MB patients ≥20 year of age.The AAIR was slightly higher in males than that in females(0.07 vs 0.05 per 100,000,P = 0.01).Compared to blacks,the incidence was nearly doubled in individuals of white ethnicity(0.06 vs 0.03 per 100,000,P< 0.001).Incidence was also lower among patients ≥40 year of age,compared to those 20-39 year of age(0.02 vs 0.11 per 100,000,P< 0.001).The white had higher 5-year and 10-year RSR than blacks.Women(<60 year of age)had higher5-year and 10-year RSR than men of the same age.Multivariate Cox analysis indicated that age(P< 0.001),surgical resection(P = 0.002)and histology(P = 0.045)were independent prognostic factors for OS in adult MB.Subgroup analyses were performed to stratify patients with vs without gross total resection(GTR),which shows that patients without GTR could benefit from radiotherapy and chemotherapy(P < 0.050).PSM revealed that age(P< 0.001),surgical resection(P = 0.012),histology(P = 0.031),insurance status(P = 0.014),and marital status(P = 0.040)had significant effects on OS.Those factors were selected to established the Nomograms for estimating 3-,5-,and 10-year OS.The C-index for the OS Nomogram was 0.746.The C-index for validation cohort was 0.683.Conclusion:This study demonstrated that adult medulloblastoma patients could benefit from GTR,young age(20-39 year of age)and histology.Our predictive Nomogram enables accurate prediction of the survival of adult MB.As for patients without GTR,radiotherapy and chemotherapy may improve the long-term survival.In addition,race-,gender-and age-disparities were linked with the trends in medulloblastoma incidence and survival.