Introduction

Australia has collected high quality data on cancer incidence, mortality and survival over many years through the efforts of state and territory cancer registries, and the Australian Institute of Health and Welfare (AIHW). Cancer Australia has now published updated national data on childhood (paediatric) cancer stage at diagnosis for 16 major childhood cancer types diagnosed in the 10-year period 2006–2015, for children aged 0-14 years. The initial release of these data in 2018 represented the first time in the world that national level data on stage at diagnosis and survival was reported for these childhood cancers. These data were collected under the Australian Government’s Investing in Medical Research - Fighting Childhood Cancer measure, from a world first trial of the Toronto Paediatric Cancer Stage Guidelines to determine stage at diagnosis for childhood cancers¹. The Guidelines and associated Business Rules used to collect these data are now endorsed internationally^2,3.

The lack of nationally consistent childhood cancer staging data has been an identified gap in Australia and internationally. Knowing the distribution of stage at diagnosis and outcomes by stage is crucial to understanding variations in survival, and can help inform further research and targeted cancer control strategies to reduce the proportion of cancers diagnosed at an advanced stage.

The 16 cancer types for which data are reported represent approximately three-quarters of all childhood cancers diagnosed in Australia and other high-income countries¹ – the relative proportions of all incident childhood cancers diagnosed in 2006–2015 are summarised in Figure 1. For each of the 16 major childhood cancer types, stage at diagnosis is reported using a tiered system as follows:

• Tier 1: A less-detailed staging system for registries with limited resources; and
• Tier 2: A more-detailed staging system for well-resourced registries. Tier 2 stage categories can be collapsed to the Tier 1 categories to enable comparisons of data from different registries.

Some cancer types have the same stage categories for both the Tier 1 and Tier 2 criteria. Where there are differences in the Tier 1 and Tier 2 stage categories, the results for these Tiers are presented separately in the text with common groupings of ‘limited’ and ‘advanced’ stage  provided, where relevant.

The childhood cancer stage data and five-year relative survival data reported for these measures were collected and analysed by the Cancer Council Queensland, with the involvement of all state and territory population-based cancer registries and major paediatric hospitals, as part of a project funded by Cancer Australia.

A detailed examination of the data, including information about the collection methodology, data sources, and guidance for interpretation are available in the following measures:

• Distribution of childhood cancer stage

• Five-year relative survival by stage at diagnosis for childhood cancers

The 16 childhood cancer types reported on the NCCI website are as follows:

*Groupings have been adapted from the International Classification of Childhood Cancers, 3rd edition (ICCC-3) Diagnostic Groups

For more information about childhood cancer, please refer to Cancer Australia’s Children’s Cancer website. More data on childhood cancer in Australia can be found at the Cancer Council Queensland’s Australian Childhood Cancer Statistics Online website.

The results provided below are for the Tier 1 staging system as described above. Data for the Tier 1 and Tier 2 staging systems are presented on each of the individual measure pages.

A high proportion of childhood cancers were staged and most cancers were diagnosed at a ‘limited’ stage

Staging completeness was high overall for each of the 16 major childhood cancers

A high proportion of cases for the 16 major childhood cancer types could be staged overall (up to 95%). The proportion of cases staged for each cancer type ranged from 83% (for acute myeloid leukaemia) to 98% (for retinoblastoma).

Of the 16 cancer types, acute lymphoid leukaemia was the most commonly diagnosed childhood cancer

Acute lymphoid leukaemia was the most commonly diagnosed childhood cancer, representing 25% of all childhood cancer cases diagnosed in the period 2006−2015.

A high proportion of cases were diagnosed at a limited stage for most of the major childhood cancer types

Of the 14 cancer types where stage can be categorised as ‘limited’ or ‘advanced’, for 12 cancer types the majority of cases were diagnosed at a ‘limited’ stage before the cancer had spread to other parts of the body, ranging from 64% for hepatoblastoma to 97% for retinoblastoma.

For Hodgkin lymphoma and neuroblastoma, a relatively high proportion of cases were diagnosed at an advanced stage

Advanced stage cancers accounted for 45% of cases for Hodgkin lymphoma and 55% of cases for neuroblastoma.

For childhood leaukaemias, a majority of cases had no central nervous system involvement

Most children diagnosed with acute lymphoid leukaemia (86%) or acute myeloid leukaemia (55%) were classified as having no central nervous system involvement (CNS-).

Childhood cancers diagnosed at a ‘limited’ stage generally had higher relative survival rates than those diagnosed at an ‘advanced’ stage

For four cancer types, five-year relative survival for advanced cancers was not examined due to small numbers

Five-year relative survival was not examined for advanced stage ependymoma, retinoblastoma, testicular cancer and ovarian cancer.

For nine cancer types, five-year relative survival was higher if the cancer was diagnosed at a limited stage, or before the cancer had spread to other parts of the body (including CNS-)

For acute lymphoblastic leukaemia, Ewing sarcoma, hepatoblastoma, medulloblastoma and other CNS embryonal tumours, neuroblastoma, non-rhabdomyosarcoma soft tissue sarcoma, osteosarcoma, rhabdomyosarcoma, and Wilms tumour, children diagnosed at a limited stage had statistically significant higher survival than those diagnosed at an advanced stage.

For non-Hodgkin lymphoma, survival was higher for children with limited (91%) compared to advanced (85%) stage disease, but these differences were not statistically significant.

Five-year relative survival was similar regardless of stage at diagnosis for acute myeloid leukaemia

There was no statistically significant difference in five-year relative survival by stage at diagnosis for acute myeloid leukaemia (77% for CNS-, and 78% for CNS+).

Five-year relative survival is high for Hodgkin lymphoma regardless of stage at diagnosis

For children diagnosed with Hodgkin lymphoma, survival was high (at least 97%) regardless of stage at diagnosis.

Some cancer types had notably lower survival outcomes when the cancer was diagnosed at an advanced stage (metastatic)

Five-year relative survival was lower than 50% for children diagnosed with advanced (metastatic) medulloblastoma and other CNS embryonal tumours (42%), non-rhabdomyosarcoma soft tissue sarcoma (34%), and osteosarcoma (36%).

Figure 1: Proportional incidence of childhood cancers, and relative stage distribution and five-year relative survival by stage at diagnosis for the three most common childhood cancer types, 2006–2015

Note: Click on the image above to expand

References

1. Aitken, J.F., Youlden, D.R., Moore, A.S., Baade, P.D., Ward, L.J., Thursfield, V.J., Valery, P.C., Green, A.C., Gupta, S., and Frazier, L.A. 2018. Assessing the feasibility and validity of the Toronto Childhood Cancer Stage Guidelines: a population-based registry study. The Lancet Child & Adolescent Health,  2(3): p. 173-179.

2. Brierley, J., Gospodarowicz, M., and Wittekind, C. 2016. The TNM Classification of Malignant Tumours, 8th edition. Lyon, France: Union for International Cancer Control (UICC).

3. International Association of Cancer Registries 2019. Paediatric Cancer Stage Guidelines - The IACR endorses the published Toronto Childhood Cancer Stage Guidelines and associated staging rules for use by population-based cancer registries.; Available from: <http://www.iacr.com.fr/index.php?option=com_content&view=article&id=153&Itemid=657>.