Summary
Background
CONCORD is a global public health programme for long-term surveillance of population-based cancer survival. The first three cycles of this programme focused primarily on adults. In CONCORD-4, for the first time, we also included all cancers in children. The WHO Global Initiative for Childhood Cancer (GICC), published in 2018, set a target for 5-year survival for all childhood cancers combined, worldwide, to reach 60% by 2030. We designed the protocol for CONCORD-4 to assess progress towards this target in as many countries as possible.
Methods
We identified population-based cancer registries from the members of the International Association of Cancer Registries and other sources. We invited 513 registries in 101 countries to submit anonymised individual records for all children (aged 0–14 years) living in their territory who were diagnosed with any form of cancer during the 30-year period 1990–2019, or later years. The data included demographic variables, the morphological type and anatomical location of the tumour, and the follow-up for the vital status of each child. We used the data for 2010–19 to construct a set of weights that reflect the global frequency distribution of childhood cancers, by age, sex, and subtype, both for the 12 major groups in the third edition of the International Classification of Childhood Cancer (ICCC-3) and for the six WHO tracer cancers prioritised in the GICC. We estimated 5-year net survival for children diagnosed during 1990–2019 by age, sex, and type of cancer, using the Pohar Perme estimator. We then used the weights to construct a Cancer Survival Index (CSI) as a weighted average of these survival estimates, for each country and each 5-year period during 1990–2019 for the 12 ICCC-3 groups and separately for the six WHO tracer cancers.
Findings
We received 679 776 individual records for children diagnosed with cancer during 1990–2022 from 307 population-based cancer registries in 68 countries and territories, 52 with 100% national coverage. We produced two sets of weights, by age, sex, and type of cancer, reflecting the global distribution of cancer in children, both for all childhood cancers and for the six WHO tracer cancers. We restricted survival analyses to 613 021 children diagnosed during 1990–2019. The 5-year CSI for all childhood cancers combined increased in most countries between 1990 and 2019. For children diagnosed during 2015–19, the CSI was more than 80% in most high-income countries, in the range 60–80% in most upper-middle-income countries, and in the range 50–60% in the five participating lower-middle-income countries.
Interpretation
The new CSI enables quantitative international comparison of trends in survival for all childhood cancers combined and for the six WHO tracer cancers, through a simple three-way standardisation by age, sex and subtype. The CSI should be a useful tool to monitor future trends. In most high-income, upper-middle-income, and lower-middle-income countries participating in CONCORD-4, the all-cancers CSI was either close to or had already passed the GICC target to reach 60% 5-year survival for all childhood cancers combined, worldwide, by 2030. The GICC target therefore may not be ambitious enough.
Funding
Cancer Research UK, Institut National du Cancer (France), St Jude Children's Research Hospital (USA), US National Cancer Institute, and Dell Technologies.
Research in context
Evidence before this study
Survival differs widely between the various types of cancer in children, and between countries defined by their World Bank national income group. In 2018, WHO published the Global Initiative for Childhood Cancer (GICC), with the central target of reaching 60% survival (presumed to be 5 years) for all childhood cancers combined, worldwide, by 2030.
No single metric exists to enable monitoring of progress towards this target. We searched PubMed for articles published in English, without date limits, using the following search string: “Population-based cancer regist*”[tiab] OR population-based registr* OR “population-based study”[tiab:~0] OR “EUROCARE”[tiab] AND “case-mix-standardised survival”[tiab:~0] OR “all cancers combined survival”[tiab:~0] OR “case-mix by cancer “[tiab:~0] OR “cancer survival index”[tiab:~0] OR “one-number index”[tiab:~0] OR “all cancers survival”[tiab:~0] OR “patient survival for all cancers combined”[ti:~0]”.
At present, the only attempt to evaluate progress towards the GICC target is derived from simulation-based model estimates of net survival for 197 countries. In most of these countries, real-world data from population-based cancer registries are not available. The estimates include all types of cancer combined in the age range 0–14 years in a single pool, despite the well known differences in survival by age, sex, and type of cancer. Some large comparisons of survival for all cancers combined were produced by the EUROCARE project, in Europe, or by NORDCAN, in northern Europe only, or in single countries (Canada, China, Denmark, and the USA).
In these studies, an estimate of survival for all cancers combined, in adults or in children, was based on a double standardisation, starting from the usual standardisation by age, followed by a further standardisation by case-mix or cancer type, and sex. This approach implies the use of two sets of weights, one for age-standardisation and another to reflect the cancer type and sex distribution of the patients included in each study. These distributions are not representative of the global population of cancer patients—in this context, children. The cancer survival indices derived for all these studies are not directly comparable either between countries or over time.
Added value of this study
The current cycle of the CONCORD programme for global surveillance of trends in population-based cancer survival (CONCORD-4) has extended coverage to include data for adults diagnosed with one of 22 malignancies, and for the first time, also includes data on all children diagnosed with a cancer during 1990–2022.
CONCORD-4 provides the largest global real-world database on childhood cancer, including data from 307 population-based cancer registries in 68 countries, 52 with 100% national coverage. We created two sets of weights that reflect the global frequency distribution of childhood cancers by age, sex, and subtype: one set for the 12 major groups defined in the 3rd edition of the International Classification of Childhood Cancer, and another for the six tracer cancers prioritised by WHO in 2021.
For each country and 5-year calendar period during the three decades 1990–2019, we then constructed a Cancer Survival Index (CSI), which enables quantitative comparisons of net survival for all childhood cancers combined, between countries and over time. Both sets of weights, which allow for a simple three-way standardisation by age, sex, and subtype, are now available for national and international research on childhood cancer survival. 30-year trends in the CSI offer a robust, long-term baseline against which to evaluate progress towards the GICC 2030 target.
Implications of all the available evidence
The CSI will facilitate monitoring of real-world progress towards the GICC target for childhood cancer survival. The CSI that includes all childhood cancers is a better indicator than the CSI based on the six WHO tracer cancers, especially for lower-middle-income countries, where diagnostic facilities are often inadequate, and the need to improve survival is even more urgent. WHO should devote even greater efforts to increase the coverage of population-based cancer registries worldwide and to facilitate data sharing for international research. In most high-income and upper-middle-income countries, impressive trends in survival for all childhood cancers combined since 1990 have already exceeded the GICC target for 2030, suggesting that a more ambitious target could be set. In low-income countries and lower-middle-income countries, where 60% of the world's children live, late presentation, abandonment of treatment, and suboptimal health-care systems are major contributors to poor survival.
Source:
Link: https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(26)00189-3/fulltext?rss=yes
____
