Background:
Central nervous system (CNS) tumours account for approximately 20% of all childhood cancers, comprising the most common solid tumours and the second most common malignancies among children after leukaemia. Up to 40% of these tumours are non-malignant (nmCNS) but they require similar therapeutic interventions as malignant tumours due to associated co-morbidity, decreased quality of life and potentially fatal outcome. There are variations in reporting of nmCNS tumours by population-based cancer registries which limits comparability of CNS tumours burden in childhood population across the world.
Objective:
To analyse the impact of registration practice, specifically registration of nmCNS tumours, on observed incidence rates.
Methods:
Tumours classified as intracranial and intraspinal according to the International Classification of Childhood Cancer were extracted from database of the International Incidence of Childhood Cancer, Volume 3. Age-standardised incidence rates per million (ASR) and 95% confidence intervals (CI) were calculated for pooled groups of registries, stratified by CNS tumour registration and coding practices and by Human Development Index (HDI) level. A questionnaire survey was conducted to explore the reasons for differences in registration of nmCNS.
Results:
Overall, 188 population-based cancer registries operating in 82 countries and territories over variable periods between 1982 and 2015 were included. A total of 113,539 childhood CNS tumours incident in children aged 0-14 years were included, yielding an overall ASR of 29.9 per million (95% CI: 27.0–32.9). Among 60 registries registering only malignant CNS tumours, the pooled ASR was 18.5 (95% CI: 13.7–23.3), compared with 33.3 (95% CI: 29.8–36.8) among 128 registries with systematic registration of nmCNS tumours. Of 100 registries located in very high HDI countries, 82 registered nmCNS tumours, whereas among 24 registries in low- and medium-HDI countries, only 9 (37.5%) did so.
The dataset included 18,527 cases of pilocytic astrocytoma (PA), with a pooled ASR of 5.3 (95% CI: 3.9–6.6). Total CNS incidence varied substantially according to PA registration. Among 39 registries excluding PA, the pooled ASR was 13.0 (95% CI: 8.4–17.6); among 103 registries classifying PA as non-malignant, the ASR was 31.7 (95% CI: 26.3–37.1); among 26 registries classifying PA as malignant, the ASR was 36.0 (95% CI: 30.4–41.5); and among 20 registries coding PA with either behaviour, the ASR was 32.3 (95% CI: 25.0–39.5). Registration of the nmCNS tumours was most commonly determined by health policy requiring registration of malignant CNS tumours only.
Conclusion/Implications:
Excluding nmCNS tumours from registration implies underestimation of CNS tumours incidence and overall cancer burden in children. The sporadic registration of nmCNS tumours in countries with lower HDI levels may reflect limitations in diagnostic capacity, difficulty to access medical records or gaps in information system. These findings, generated through the ChildGICR collaboration, underscore the critical importance of comprehensive registration of mnCNS tumours in childhood cancer surveillance.