EPID-03. CURE FRACTION AND LONG TERM SURVIVAL ESTIMATES FOR PEDIATRIC CENTRAL NERVOUS SYSTEM TUMORS IN EUROPE – RESULTS OF THE POPULATION-BASED EUROCARE PROJECT

Abstract

Abstract BACKGROUND Time trends in cure fractions and up-to-date long-term survival have not been reported for clinically relevant groups of pediatric central nervous system (CNS) tumors although it provides important information for planning long-term clinical follow-up. METHODS We analyzed data on 15,242 children (<15 years) diagnosed with a CNS tumor between 1998-2013 from 31 European countries with follow-up until 31/12/2014. Cure fraction, the proportion of patients not anymore at risk of dying from progression or relapse was estimated using a mixture cure model with a parametric cancer survival curve function and assuming constant long-term mortality from other causes, including other cancers. Additionally, model-based up-to-date 15-year observed survival (OS) was estimated for the period 2010-2013. RESULTS The cure fraction for ependymomas increased from 65% in 1998-2001 to 79% in 2010-2013. However, as a result of the high long-term mortality rate of 16 per 1000 surviving patients per year, ependymomas had an estimated 15-year OS of 62%. For diffuse astrocytomas the cure fraction increased 6 percent-points to 84%, and 15-year OS was 74%. Cure fraction for glioblastomas increased from 15% to 20%. Long-term mortality for surviving glioblastoma patients was null resulting in a comparable projected 15-year OS of 20%. For medulloblastomas the cure fraction increased with 4 percent-points to 56% in 2010-2013, and a comparable 15-year OS of 56%. AT/RTs including PNETs together with high-grade gliomas (HGGs) were the only groups for which the cure fraction did not change over time: 34% and 24%, respectively. Estimated 15-year OS was 28% for AT/RTs including PNETs and 19% for HGGs. CONCLUSIONS As a result of high long-term mortality, the difference between cure fraction and 15-year OS was the highest for ependymomas and diffuse astrocytomas. Further investigation into underlying reasons (e.g., late effects of treatment) are needed to reduce the extra risk of deaths.