Early diagnosis key to improving childhood cancer survival

A major study by UCL and Fondazione IRCCS Istituto Nazionale dei Tumori in Milan (INT) researchers has for the first time shown in detail how far children’s cancer has spread at diagnosis in a way that can be compared between countries.

While poorer survival following late-stage diagnosis is well recognised, the study is the first to show that differences in tumour stage at diagnosis may explain why childhood cancer survival varies between some European regions and tumour types.

Published in JAMA Network Open, the researchers analysed high-quality data from the cases of 9,883 children from 73 population-based cancer registries across 27 countries, mainly European, diagnosed with six different cancers – neuroblastoma, Wilms’ tumour, medulloblastoma, osteosarcoma, Ewing sarcoma and rhabdomyosarcoma – between 2014 and 2017.

They found that across all six cancers, survival at three years was strongly linked to tumour stage at diagnosis, with survival decreasing as stage increased.

Using Central Europe* as the benchmark and looking at the data by region, they found significant variation in overall three-year survival rate between some regions in the cases of four of those cancers.

In the case of neuroblastoma – tumour of adrenal gland and some other sites – they found that children had lower survival rates in the UK and Ireland compared with Central Europe, a difference that could be explained by the relatively later stage at which that cancer was diagnosed in the UK and Ireland.

In the case of Ewing sarcoma (bone tumour), by contrast, in Eastern Europe and in the UK and Ireland, the lower survival rate compared with Central Europe wasn’t accounted for by later diagnosis. In fact, the difference was only seen for patients whose tumour had spread. The researchers suggested that other factors such as where the tumour had spread to and any differences in treatment approaches for children whose cancer has spread need to be investigated.

Until now, the lack of reliable data on tumour stage at diagnosis has meant it was impossible to use registry data to shed light on the difference in child cancer survival rates by country.

The UCL and INT researchers’ International Benchmarking of Childhood Cancer Survival by Stage (BENCHISTA) project has improved the way national cancer registries collect childhood cancer data so it can be compared reliably.

Through an agreement which took 18 months to broker, the countries involved in the study – 23 European countries including the UK but also Brazil, Canada, Australia and Japan – improved their data collection techniques to ensure at least 90% of the child cancer cases in their countries had tumour stage recorded (a level deemed acceptable for accurate international data analysis) and in a consistent way (using what is known as the Toronto Childhood Cancer Stage Guidelines, which were developed in the 2010s).

The result is that for the first time researchers are in a position to better understand why childhood cancer survival rates differ from region to region with a view to guiding efforts to improve those rates.

Their findings highlight the need to examine other factors, such as differences in treatment approaches, access to specialised care, and health system differences.

This will be a focus of the next phase of the BENCHISTA project, which will also report five-year survival, the standard epidemiology cancer outcome measure.

Joint senior author Professor Kathy Pritchard-Jones (UCL Great Ormond Street Institute of Child Health) said:

By analysing population-level data from cancer registries across many countries, we have been able to better understand why childhood cancer survival still differs internationally.

Our findings show that diagnosing cancer earlier and accurately assessing how far it has spread can make a meaningful difference to survival for many children. At the same time, the study highlights that early diagnosis alone will not address all disparities, and further work is needed to understand and tackle other contributing factors.

We have, for the first time, provided unbiased, population-level evidence for later diagnosis of some childhood cancers in the UK and Ireland.

Altogether, the findings from BENCHISTA will provide unbiased evidence regarding the need for both cancer-specific strategies and broader improvements in diagnosis and care, the researchers say.

Corresponding author Dr Laura Botta, a statistician in the INT’s Evaluative Epidemiology Unit, said:

Our project not only emphasises the importance of gathering high-quality data and strengthening cooperation between clinicians and population-based cancer registries but also serves as a key tool to uncover the reasons for survival differences between countries for children with cancer.

Our results provide the first population-based and international evidence that, in some countries, paediatric tumours are diagnosed at more advanced stages compared to a group of European countries with the most consistent practices, while in others, staging procedures do not meet recommended standards – clearly impacting prognosis. We hope these findings will help guide health policies that are vital to saving lives.”

The study was funded by Children with Cancer UK and Associazione Italiana per la Ricerca sul Cancro (AIRC).

Gavin Maggs, Chief Executive of Children with Cancer UK, said:

We are proud to have funded this important study, which provides powerful new evidence on the role that early diagnosis plays in improving childhood cancer survival.

By bringing together experts from around the world, BENCHISTA shows how internationally comparable data can guide global efforts to ensure children are diagnosed as early as possible and receive the right care at the right time. Cancer does not recognise borders, and neither should the chances of surviving it.

Ashley Ball-Gamble, chief executive of CCLG: The Children & Young People’s Cancer Association, said:

This study confirms that, for some childhood cancers, diagnosis in the UK still takes longer than it should.

Because delays can affect a child’s chance of survival, campaigning for faster recognition is more important than ever.

Our Child Cancer Smart campaign aims to tackle these delays by giving GPs the tools to identify cancer sooner – and by helping parents understand the warning signs.

Angela Polanco, the project’s patient and public involvement lead, whose daughter Bethany was diagnosed with Wilms tumour, said: 

This study provides clear evidence that we need to do more to ensure children affected by cancer have access to timely and accurate diagnosis, appropriate first-line treatment, and specialist care, wherever they live.

Collecting and analysing this information at a population level is a crucial step towards reducing inequalities and improving survival for children with cancer.

Rhabdomyosarcoma is a muscle tumour, medulloblastoma is a brain tumour, osteosarcoma – like Ewing sarcoma – is a bone tumour, and Wilms’ tumour is a kidney tumour.

Study limitations

As with many rare disease studies, the relatively small number of cases and the three-year follow-up period mean that some longer-term survival differences may not yet be fully captured.

* In the study, the region of Central Europe was made up of Austria, Belgium, France, Germany, Switzerland and the Netherlands.