Survival from Brain Tumours in Children: A Systematic Review of Geographical Variation and Time Trends

Abstract

Background: Despite their rarity, brain tumours are among the leading causes of cancer-related deaths in children. Globally, five-year survival for all-brain tumours combined varies widely. Morphology drives treatment, and hence health services allocation. However, it is currently unknown how survival by brain tumour sub-type varies between countries and whether outcomes have improved over time. Methods: In this systematic review, we considered longitudinal, observational studies comprising children diagnosed with a primary brain tumour. We searched six electronic databases from database inception to 30 September 2018. The outcome of interest was five-year survival estimated through a time-to event analysis. We extracted data from published reports. The systematic review is registered with PROSPERO, number CRD42018111981. Findings: We identified 5,244 potentially eligible records. Forty-seven papers were included. Only three studies were conducted in low-income or middle-income countries. Five-year survival from embryonal tumours rose from 37% in 1980 to around 60% in 2009. While survival for medulloblastoma improved substantially (from 29% to 73% during 1959-2009), survival for PNET waved over time (1973-2009) and between countries. Five-year survival from astrocytoma as a broad morphology group barely changed during 1982-2009 (from 78% to 89%), but interpretation of trends was hampered by missing information on behaviour. Indicators of data quality were often either missing or incomplete, with the proportion of microscopically verified tumours being the most commonly reported parameter. Interpretation: An improvement in outcomes occurred for embryonal tumours, while little change was observed for astrocytic tumours. The available evidence is restricted to high-income countries. Interpretation of the literature is hindered by the heterogeneity between the studies, and by the quality of the available data. Larger international studies based on a common data collection protocol, would allow more robust comparison. Funding: Dr Fabio Girardi MD is supported by the Davidson and O'Gorman Fellowship from Children with Cancer UK. Declaration of Interest: We declare no competing interests. Ethical Approval: The systematic review is registered with PROSPERO, number CRD42018111981.